Quarterly Non-financial Accounts by Institutional Sector (QSA) in Belgium ESA 2010 Sources and Methods

Transcription

1 Quarterly Non-fnancal Accounts by Insttutonal Sector (QSA) n Belgum ESA 2010 and Thrd edton February 2016

2 2. Quarterly Non-fnancal Accounts by Insttutonal Sector and ESA 2010 Thrd edton (March 2016) BELGIUM Table of contents 1. General descrpton Organsatonal aspects General descrpton Exhaustveness and balancng Use of econometrc modellng Plausblty checks Techncal framework Consstency wth related datasets Consstency wth annual data Consstency wth man aggregates of sector S1 (QNA) Consstency wth government data (QNFAGG) Consstency wth Bop data (BPM6 verson) Consstency wth fnancal accounts Short descrpton of the methods d to estmate back data Seasonal adjustment polcy Release and revson polcy Release polcy Revson polcy Major revsons (consstency wth updated annual data) Revsons of data for the latest quarters Methodologcal revsons Remarks or problems Future plans Descrpton by transacton General remarks Overall pcture Detaled revew of transactons Output (P.1) and ntermedate consumpton (P.2) Value added, gross (B1.g) Fnal consumpton expendture (P.3) Gross captal formaton (P.5) Gross fxed captal formaton (P.51) Changes n nvento and acqustons less dsposals of valuables (P.5N) Exports (P.6) and mports (P.7) of goods and servces Compensaton of employees (D.1) Taxes on producton and mports (D.2) Taxes on products (D.21) Other taxes on producton (D.29) Subsdes (D.3)... 23

3 Subsdes on products (D.31) Other subsdes on producton (D.39) Intet (D.41) Property ncome other than ntet (D.4N) Dvdends (D.421) Wthdrawals from the ncome of quas-corporatons (D.422) Renvested earnngs on drect foregn nvestment (D.43) Property ncome attrbutable to nsurance polcy holders (D.441) +Investment ncome payable on penson enttlements (D.442) Investment ncome attrbutable to collectve nvestment fund (CIF) shareholders (D.443) Rents (D.45) Current taxes on ncome, wealth, etc. (D.5) Socal contrbutons (D.61) Actual socal contrbutons (D61x = D D D D.61SC) Imputed socal contrbutons (D.612) Socal benefts other than socal transfers n knd (D.62) Socal transfers n knd (D.63) Other current transfers (D.7) Net non-lfe nsurance premums (D.71) Non-lfe nsurance clams (D.72) Other current transfers (D.7N) Adjustment for the change n net equty of hoholds n penson funds erves (D.8) Captal transfers (D.9) Captal taxes (D.91) Investments grants and other captal transfers (D.9N) Consumpton of fxed captal (P51C) Acqustons less dsposals of non-fnancal non-produced assets (NP) Boxes Relatve mportance of pre-exstng data and approprate estmates of mssng data for nonfnancal corporatons, fnancal corporatons and hoholds... 5 Addtonal explanaton for the estmaton of quarterly profle of dvdends pad by dent enterprses Tables Table 1 - Resdual sector selected to balance QSA... 8 Table 2 - Breakdown of value added by ndustry between nsttutonal Table 3 - Breakdown of the quarterly compensaton of employees pad between nsttutonal sectors Table 4 - Quarterly profle of dvdends pad by Belgan corporatons Table 5 - Quarterly profle of dvdends pad by the t of the world accordng to natonal accounts Table 6 - Quarterly profle of dvdends receved by the t of the world accordng to natonal accounts Table 7 - Method d to estmate annual dvdends Annexes Methodology by transacton - Overall pcture... 51

4 4. 1. GENERAL DESCRIPTION 1.1. ORGANISATIONAL ASPECTS Quarterly non-fnancal sector accounts (QSA) are establshed wthn the "Natonal / regonal accounts and busness surveys" dvson of the Natonal Bank of Belgum's General Statstcs Department, on behalf of the Natonal Accounts Insttute (NAI). Annual and quarterly sector accounts are compled wthn the same dvson of the NBB, as well as the most mportant nputs such as man aggregates and government accounts. Fnancal accounts are establshed wthn the "Fnancal statstcs" dvson of the General Statstcs Department. Balance of payments (Bop) data are establshed wthn the External statstcs dvson. An organzaton chart of the General Statstcs Department s provded below. Contact persons are: Béatrce Thry (beatrce.thry@nbb.be) and Catherne Rgo (catherne.rgo@nbb.be)

5 SOURCES Two man pre-exstng quarterly data sets serve as nput when establshng QSA: The quarterly man aggregates for the whole economy (S.1) whch are publshed each quarter by the NAI, wth a delay of 60 days after the closng date of the quarter. The quarterly accounts for general government (S.13) whch are compled ndependently of the accounts for other sectors, wth a delay of 85 days after the closng date of the quarter. The quarterly man aggregates (QNA) are based on the followng short-term nformaton: o o o o ndustral producton ndces (frst verson avalable at t+40 days); VAT data (frst confdental verson avalable at t+50 days); socal securty data (frst verson avalable at t+4 months); Bop data (frst verson avalable at t+45 days); The quarterly non-fnancal accounts of general government (QNFAGG) s based on ndcators related to general government taxes and expendture : o tax revenue (avalable at t+80 days) o non-fscal and non-parafscal revenue: specfc nformaton s avalable for certan headngs (nter-muncpal body dvdends, non-recurrent exceptonal payments, etc); o ntet payments on the federal authorty s publc debt (avalable at t+60 days); o other expendture by the federal authorty, communtes and regons: trend for total expendture s avalable at t+45 days, together wth certan non-recurrent exceptonal expendture; o socal contrbutons pad to the Natonal Socal Securty Offce (NSSO) (avalable at t+80 days) Pre-exstng data are not suffcent to comple complete QSA. To estmate quarterly mssng drect nformaton, dfferent methods have been adopted dependng on the avalablty of addtonal sources of nformaton or ndcators (cf. 1.3 below). In some cases, specfc methods have been developed whch take nto account addtonal nformaton stemmng from: o balance sheets of non-fnancal companes (exhaustve admnstratve annual source); o fnancal accounts (quarterly source); o MIR data on ntet rate (monthly source) o reportng schemes of fnancal corporatons (exhaustve, partly annual, partly quarterly source); o Belgan Bop (quarterly source); o Bop of European nsttutons (quarterly source); o Natonal Socal Securty Offce data (quarterly admnstratve exhaustve source). In some other cases (when no ndcator s avalable), a smple lnear trend s defned. To guarantee the overall consstency of sector accounts, some transactons are also calculated as dual tems to balance the system throughout the nsttutonal sectors and on the bass of the dentty that, for the total of nsttutonal sectors, the ources must equal the s. As most of the pre-exstng quarterly data are already avalable for the total economy (S.1), the general government (S.13) and the t of the world (S.2), the estmate of the remanng transactons based on addtonal nformaton and specfc methods manly concerns the accounts for non-fnancal corporatons (S.11), fnancal corporatons (S.12) and hoholds 1 (S.1M). Defnng lnear trends or calculatng tems as balancng transactons occur for a set of transactons 1 As stated n the European Parlament and Councl Regulaton 1161/2005, quarterly sector accounts are bult up for hoholds (S.14) together wth NPISH's (S.15). In the pent document, the groupng of both nsttutonal sectors s mentoned wth the codfcaton S.1M.

6 6. whch are not amongst the most crtcal one, as llustrated n Box 1. Thanks to the approach d, an effcent producton of robust QSA n a short tme can be ensured. Ths also allows meetng the requrements set out n Artcle 5/2 of the European Parlament and Councl Regulaton 1161/2005 whch mposes QSA to be consstent wth the quarterly non-fnancal accounts for the general government and the quarterly man aggregates of the total economy. As stated n the same regulaton (Artcle 5/3), QSA are also always completely algned wth the corpondng annual data transmtted under the data transmsson programme of the ESA2010 Regulaton. Box - Relatve mportance of pre-exstng data and approprate estmates of mssng data for non-fnancal corporatons, fnancal corporatons and hoholds QSA are buld up by ncorporatng pre-exstng data whch have to be flled n wth approprate estmates when tems (transactons) are mssng. The relatve mportance of pre-exstng data s llustrated below for non-fnancal and fnancal corporatons separately and for hoholds. The tables show, for the account of each of these nsttutonal sectors, the number of transactons 2 and the total and average amount of these transactons whch come drectly from pre-exstng data.e. manly quarterly man aggregates for the whole economy or counterpart of quarterly accounts for general government or Bop se. If the transactons for whch a specfc estmate method usng ad hoc nformaton has been developed are consoldated wth the pre-exstng data sets, t appears that these data repent the substance of the accounts as they count for some 73 to 94% of the total amount of transactons, dependng on the nsttutonal sector 3. These transactons generally concern the man tems as ther average amount s qute large. On the other sde, the average amount of transactons whch are estmated wth a trend or as dual tem s more lmted. These transactons count for some 6% to 13% of the total amount of all the relevant transactons for hoholds and non-fnancal corporatons. They repent some 27% of the relevant flows n the case of fnancal corporatons. Components of the quarterly non-fnancal corporatons account (S.11) Pre-exstng data and specfc estmates Other (trends and dual tems) Number of relevant varables Cumulatve amount (mllons euro) 123,365 17,671 % of total 87.5% 12.5% Average amount per relevant varable (mllons euro) 8, Components of the quarterly fnancal corporatons account (S.12) Pre-exstng data and specfc estmates Other (trends and dual tems) Number of relevant varables Cumulatve amount (mllons euro) 32,063 13,201 % of total 70.8% 29.2% Average amount per relevant varable (mllons euro) 2, When methods have been defned for sub-transactons, hgher-level transactons have not been consdered as relevant transactons n the pent assessment exercse. The relevant transactons appear on the ources or on the s sde of the account. 3 Calculatons have been drawn for the four quarters of 2012 (as establshed for the January 2016 transmsson) and averaged.

7 7. Components of the quarterly hoholds account (S.1M) Pre-exstng data and specfc estmates Other (trends and dual tems) Number of relevant varables Cumulatve amount (mllons euro) 175,005 1,038 % of total 89.4% 10,6% Average amount per relevant varable (mllons euro) 9,211 1, METHODS General descrpton Three sets of transactons can be drawn accordng to ther quarterly avalablty. (A) For the followng transactons, quarterly data are avalable by nsttutonal sector: - fnal consumpton expendture (P.31 and P.32) - exports and mports of goods and servces (P.6 and P.7) - taxes and subsdes on products (D.21 and D.31) - current taxes on ncome, wealth (D.5) - socal contrbutons relatng to the State socal securty system (part of D.61) - socal benefts other than transfers n knd and relatng to the State socal securty system (part of D.62) - socal transfers n knd (D.63) - current nternatonal cooperaton (D.74) - VAT and GNI based EU own ources (D.76) - captal taxes (D.91) - complete sequence of government account. (B) The followng transactons are avalable on a quarterly bass for the total economy (S.1) but have to be broken down amongst the nsttutonal sectors: - value added (B.1.g) (avalablty by 38 ndust) - compensaton of employees (D.1) - gross captal formaton (P.5) - other taxes on producton (D.29) - other subsdes on producton (D.39) - consumpton of fxed captal (P.51c) - acqustons less dsposals of non-fnancal non-produced assets (NP). For the two frst tems of ths second group, a specfc method to break down quarterly data by nsttutonal sector has been developed: the breakdown by nsttutonal sectors of quarterly value added and quarterly gross fxed captal formaton s estmated wth reference to a corpondng annual key that s appled to quarterly data. For added value, such a breakdown by sector s calculated for each ndustry. For the other transactons, a smple smoothng method of annual fgu (Chow and Ln or Fernandez method wth a lnear temporal se as ndcator) s d together wth balancng procedu that perve consstency wth the whole economy constrants. (C) For the followng transactons, no pre-exstng quarterly nformaton s known:

8 8. o o o o o o o o ntet ncome (D.41) dstrbuted ncome from corporatons (D.42) renvested earnngs on drect foregn nvestment (D.43) other nvestment ncome (D.44) rent (D.45) tems related to nsurance transacton (part of D.61 and D.62, D.71, D.72, D.8) mscellaneous current transfers other than related to S.13 (D.75) captal transfers other than related to S.13 (D.92+D.99) A specfc method to buld up quarterly data has been developed for the followng transactons: D.41, D.42, D.43, D.443, by collectng nformaton comng from the balance sheets of non-fnancal companes, the reportng schemes of fnancal corporatons, fnancal accounts and the Bop data. In the other cases, a smple smoothng method of annual fgu s d (Chow and Ln or Fernandez method wth a lnear temporal se as ndcator) wth due regard for accountng constrants Exhaustveness and balancng The methodology ensu exhaustveness of the whole dataset. Procedu are systematcally appled to guarantee that: The sum of the data for the four quarters of a year s equal to the annual data : Q Y Ths s de facto the case when usng pre-exstng quarterly data whch are consstent wth annual data. (In September, where new released annual data have not been converted yet n new offcal quarterly data that must be ncorporated wthn the QSA as an external nput, a procedure for the specfc purpose of QSA s proceed to generate new quarterly se that are algned wth the new annual data. Ths procedure generates a new quarterly tme se algned on the new annual one by usng the old quarterly tme se as an ndcator.) The smoothng method mentoned above also guarantees that the sum of quarters equals the annual data. Accountng constrants are met (horzontal consstency). (a) for each transacton, quarterly data for the whole economy are equal to the sum of quarterly data for the dfferent domestc nsttutonal sectors: S.1 S.11 S.12 S.13 S.1M and (b) for each dstrbutve transacton, total ources equal total s: TOT_(S.1 S.2) TOT_(S.1 S.2) Whether the total s calculated on the ources sde or on the s sde of an account depends for each transacton on the nformaton avalable on each sde. The total s calculated on one sde, as the sum of the data for the domestc sectors and the t of the world. It s then d as a constrant on the other sde of the account and an nsttutonal sector s chosen to be the

9 9. dual sector whch balances the system. The "dual sector" s ether S.11 or S.1M, dependng on the transacton 4. No dual sector s defned for transactons whose quarterly pattern s bult up usng a specfc method (D.41, D.42, D.43, D.44, D.75, D.92+D.99). Transacton Table 1 - Resdual sector selected to balance QSA Total constrant (S1+S2) calculated on : Resdual sector P.52+P.53 (=P.5N) ---- S.11 D.1 s sde S.1M_ D.29 ources sde S.11_ D.39 s sde S.11_ D.45 s sde S.1M_ D.61 ources sde S.1M_ D.62 s sde S.1M_ D.63 s sde S.1M_ D.71 ources sde S.1M_ D.72 s sde S.1M_ D.8 s sde S.1M_ P51C ---- S.11 NP ---- S.11 ESA2010 balancng tems n the sequence of accounts are calculated on the bass of the upstream transactons except value added (B.2, B.5, B.6, B.8 and B.9). Value added (B.1) for the total economy s calculated as the sum of the value added of the domestc sectors Use of econometrc modellng Gven the large avalablty of quarterly nformaton and the chosen approach, there s no need for an econometrc modellng n the Belgan QSA. In the cases of mssng quarterly data.e. when there s no pre-exstng quarterly data, when no specfc methodology has been developed, when a transacton cannot be estmated as a dual tem to balance the system and when there s no approprate ndcator, quarterly data are estmated by smoothng out the corpondng annual fgure over the four quarters of the year, by applyng a lnear trend 5. The smoothng out of annual data s obtaned by usng the Chow and Ln (or Fernandez) method wth a lnear temporal se as ndcator. The choce of a smooth dstrbuton s not only justfed by the unavalablty of adequate quarterly ndcators but also by the lmted annual amounts nvolved or by the fact that the transacton concerned by the smoothng method can reasonably be consdered as qute stable from quarter to quarter. If the smoothng method generates quarterly negatve values, manual adjustments are ntroduced, generally by draggng, wthn a same year, the negatve value to the prevous or next quarter that dsplays a postve value. Ths procedure apples to a very few se of lmted value. When annual accounts are not yet avalable, quarterly data for the transactons concerned by the smoothng method are bult up by extendng the lnear trend. In order to smooth data, as for seasonal adjustment (see 1.6), software JDemetra+ s d. 4 Ths depends on the "expert judgment". The greater the relatve sze of one sector for a specfc transacton s, the better t s to be chosen as a dual tem. It can then absorb the changes n the data mputed to the others sectors wthout beng too much affected (n relatve terms) and wth a lmted rsk to become negatve. 5 The annual amounts are not dvded by 4 n order to avod an artfcal break between the fourth quarter and the frst quarter of the subsequent year.

10 Plausblty checks Data s montored as regards nternal consstency and consstency wth other datasets (ASA, QNA, QNFAGG). There s also a regular revson analyss. Outlers are montored, especally n the government account. Metadata s provded each quarter to Eurostat, accordng to the agreed template among TF-QSA members Techncal framework QSA are manly establshed on the bass of Excel sheets. Quarterly dsaggregaton of annual data (on the bass of ndcators or lnear) are computed wth the help of JDemetra+. Seasonal and workng days adjustments are computed wth JDemetra CONSISTENCY WITH RELATED DATASETS Consstency wth annual data By constructon, QSA data are always consstent wth corpondng annual data. The same consoldaton rules apply to all nvolved data sets (ASA, QSA, QNFAGG) Consstency wth man aggregates of sector S1 (QNA) Some man aggregates transmtted to Eurostat n Table 1 (60 days after the end of the quarter) are ntegrated nto QSA for the whole perod of transmsson. Ths concerns the followng transactons: o o o o o o fnal consumpton expendture of hoholds and NPISH's (P.3_S.1M); exports of goods and servces (P.6_S.1); mports of goods and servces (P.7_S.1); consumpton of fxed captal (P51C_S.1); compensaton of employees (D.1_S.1) added value for all sectors except S.13. For other tems, dscrepances between data transmtted n Table 1 and QSA mght occur. Ths relates wth the establshment of prortes n the avalable sources of data when QSA are compled: general government data transmtted 85 days after the end of the quarter, always have prorty over the data transmtted n Table 1. Ths apples, for nstance, to the followng tems: o o o fnal consumpton expendture by the general government (P.3_S.13); gross fxed captal formaton by the general government (P.51_S.13); taxes less subsdes on products (D.21-D.31). In these cases, as QSA ncorporate more recent general government nformaton (whch s transmtted some 25 days after Table 1 and ncorporates updates for government transactons), dscrepances between Table 1 and QSA occur generally for the latest quarters 6. As QSA refer to the most recent nformaton for consumpton and nvestments of the general government, the dfferences wth fgu transmtted n Table 1 are offset by adjustng the amount of changes n nvento for the total economy (P.52_S1) whch therefore dffer from the one 6 Government data for former quarters are ndeed generally stable and have not been revsed, when complng QSA, snce the latest transmsson of Table 1. More extended dscrepances (for the whole perod of transmsson) could however occur n the case of revsed annual (hence quarterly) general government data that are not yet ncorporated n the man aggregates (n March and September). In ths specfc case, prorty s also gven to the newly released government data.

11 11. transmtted n Table 1. Ths adjustment helps to keep consstency wth quarterly estmates on the producton sde of the economy and to perve quarterly GDP transmtted n Table 1. However, quarterly GDP can fnally dffer from the one transmtted n Table 1 beca of possble dscrepances n taxes less subsdes (usng general government data when complng QSA 7 ) Consstency wth government data (QNFAGG) As mentoned above, QSA nclude the quarterly non-fnancal accounts for general government. These are ntegrated as such wthout any correcton. Ths mples that QSA can only be bult up when government data are fnalzed. Those data always have prorty over data transmtted n Table 1 and hence can nterfere wth them (cf above) Consstency wth Bop data (BPM6 verson) 8 Recently, at annual level, some progs has been made n the feld of consstency between NA and Bop data, especally for tme span For transactons D.1, D.5, D.6, D8, D.9, NP, as well as for D.2, D.3 and D.7 vs-à-vs general government, consstency s fully met. Some dscrepances reman for D.4 and D.7, but should dsappear at medum notce. As regards exports (P.6) and mports (P.7) of goods and servces, some dscrepances wll reman, as consequence of balancng procedure of the NA man aggregates. However, the sze of these dscrepances s lmted compared to the huge gross amounts of P.6 and P.7. At quarterly level, the same stuaton prevals. However, for the more recent quarters some nfraannual dscrepances may appear for all transactons, due to vntage Consstency wth fnancal accounts The general rule s that statstcal dscrepances between net lendng/borrowng accordng to fnancal accounts on the one hand (B.9F) and accordng to non-fnancal accounts (B.9) on the other hand are not hdden. The nconsstences are the most mportant for hoholds (S.1M) and non-fnancal corporatons (S.11). Specfc rules are defned n fnancal accounts for S.2: the net lendng/borrowng n the fnancal accounts s by constructon set equal to the one that s determned n the non-fnancal accounts and the adjustment s mputed n other accounts recevable/payable. By constructon, there s no statstcal dscrepancy for the t of the world SHORT DESCRIPTION OF THE METHODS USED TO ESTIMATE BACK DATA In general, the whole data set, from 1999 on, s estmated accordng to the same methodology. However, some exceptons occur: For ntet (D.41), for the quarters of the years before 2003 quarterly matrces were not bult up beca ths work proved to be very tme-consumng. A smplfed ad hoc procedure, whch reles on smoothng trends, s appled to get quarterly estmates of total ntet receved and pad by each nsttutonal sector wthout nformaton on the counterpart sector (see pont 2.3.9). 7 Dfferences n the value added of the general government are counterbalanced wth an equvalent adjustment (wth the opposte sgn) n the value added of non-fnancal companes. Ths helps to lmt the dvergence between the GDP released n the man aggregates and the one whch s consdered n the QSA. 8 Consstency s not ensured for Bop data establshed on the bass of BPM5 ( ).

12 12. For nvestment ncome attrbutable to collectve nvestment fund (CIF) shareholders (D.443) for the quarters of the years before 2009, t was not possble to mplement the detaled quarterly calculaton method, as Bop-BPM6 data s not avalable before 2009 and as the natonal accounts estmaton s establshed on annual bass. Consequently, a smplfed methodology s d,.e. the annual fgure for each sector (ource or ) s spread over quarters usng a key. Ths key s equal to 2/3 of the corpondng key for ntet and 1/3 of the corpondng key for dvdends. A fnal adjustment s appled to ensure that total ources are equal to total s (see pont ). For dvdends (D.421), the quarterly dsaggregaton of annual data s based on fxed allocaton keys for years before SEASONAL ADJUSTMENT POLICY Seasonal adjustment s calculated usng JDemetra+ software. Se are also adjusted for calendar days when relevant. Seasonally adjusted se comng from QNA or QNFAGG are taken on board as such n QSA. All other QSA se (except those calculated by smoothng) are tested. Se showng seasonalty are seasonally and calendar days adjusted usng JDemetra+. All se are tested for seasonalty once a year and seasonalty models are also revsed once a year, n September. In order to ensure horzontal consstency of the data set by transacton, the same accountng constrants as explaned at pont are appled to seasonally adjusted data set RELEASE AND REVISION POLICY Release polcy Snce Aprl 2007, the quarterly accounts for general government (S.13) are offcally released 9. The accounts for general government are publshed at t+85 days on the onlne database of the Central Bank of Belgum for the December and June releases. For the second and fourth quarters, due to EDP procedure, the data are publshed around md-aprl and md-october (between t+110 and t+120 days). Some days later on, a quarterly publcaton of the whole QSA s released. Ths release ncludes an onlne publcaton of the whole dataset, except sub-catego of D.4 and D.7, and key ndcators. The key ndcators selected are the same as the one selected by Eurostat when releasng the European sector accounts. At the same tme, a paper-publcaton "Comptes natonaux-comptes trmestrels" "Natonale rekenngen-kwartale rekenngen" s released, together wth a ps release, gvng some comments on the key ndcators. The publcaton of QSA data and ps release follows the schedule below: o yyyyq1 t +100 days o yyyyq2 t +120 days (benchmarkng on annual data + EDP) o yyyyq3 t +100 days o yyyyq4 t +120 days (EDP) Lnk: 9 The accounts for general government are publshed accordng to 'Short-Term Publc Fnance Statstcs' (STPFS) synoptc tables (recepts, expenses and balance wth some detals), not accordng to the full QSA sequence of accounts.

13 Revson polcy Major revsons (consstency wth updated annual data) When new annual natonal accounts are released, QSA are updated to be consstent wth them. In ths case, QSA can be revsed for several years backwards. Ths occurs when producng QSA at the end of September: QSA are then updated to be algned on the new set of ASA also released n September (ncludng new government accounts). QSA are always consstent wth revsed quarterly data for the general government. Any updatng of these data for general government s ncorporated nto QSA, even f ths updatng nvolves not beng consstent any longer wth other transmsson programs (notably Table 1) Revsons of data for the latest quarters When annual natonal accounts are not yet avalable, quarterly data can be revsed when a new quarter s estmated. The revson of QSA manly depends on the revsons brought to pre-exstng data for the latest quarters Methodologcal revsons Methodologcal mprovements can be brought to QSA, up to now on the occason of any transmsson durng the year. Such revsons are mplemented for the whole perod of transmsson (backwards to 1999Q1). Methodologcal revsons to QSA never contradct other data transmsson programs REMARKS OR PROBLEMS Rents (D.41) are avalable ncludng and excludng Fsm adjustment. Snce the mplementaton of ESA2010, the same consoldaton rules apply to all nvolved data sets (ASA, QSA, QNFAGG). Fnancal and non-fnancal accounts are establshed on the bass of the same classfcaton among sectors ( of the same busness regster) FUTURE PLANS The effcent producton of robust QSA n a very short tme s ensured thanks to the approach chosen whch s easy to work out. Ths general prncple should not be departed from at the rsk of compromsng tmelness. Smoothng procedu should be, as far as possble, avoded even f these only apply n Belgum for transactons whch are lmted n sze and hence do not have a substantal effect on quarterly accounts. Explanatory works, suggested that t would be dffcult to fnd relevant quarterly ndcators for other transfers (except those related to the t of the world), rents and prvate socal securty schemes. Some more attenton could be pad n a near future to transfers related to nsurance, as some quarterly nformaton has been recently made avalable. For all se estmated wth smoothng out the annual data, the Belgan NAI wll contnue to pay due attenton to the developments that could occur n these felds to fnd out approprate quarterly ndcators. Quarterly Supply and Use tables should be establshed at medum notce. Ths wll enable the estmaton of P.1 and P.2 by nsttutonal sector and mprove the accuracy of the Belgan QSA.

14 DESCRIPTION BY TRANSACTION 2.1. GENERAL REMARKS The establshment of quarterly accounts for the general government (S.13) s descrbed n a document Précsons méthodologques relatves aux comptes non fnancers trmestrels des admnstratons publques. The document s avalable on the NBB webste ( The buldng up of the quarterly man aggregates s descrbed n the specfc Manual "Quarterly natonal accounts of Belgum - Methodologcal nventory: Descrpton of sources and methods d - December 2007". The document s avalable on the NBB webste ( These two documents are stll based on ESA1995 concepts, but the descrbed prncples and methodology reman vald. The pent descrpton does not dwell on methodologes whch are detaled n one of those two manuals. Cross-references to ths document are mentoned n the followng text OVERALL PICTURE An overall pcture for each transacton s provded n Annex 1 and Annex DETAILED REVIEW OF TRANSACTIONS Output (P.1) and ntermedate consumpton (P.2) These two optonal tems are not compled by nsttutonal sector so far on a quarterly bass and are not transmtted to Eurostat. Ths s due to the lack of quarterly Supply-Use tables that could enable P.2 to be calculated usng techncal coeffcents. At the tme beng, short-term ndcators (VAT data, ndustral producton ndces etc.) are d to drectly estmate value added. However, some mprovement n ths feld should be brought n the future (see 1.9) Value added, gross (B1.g) On the bass of short term nformaton such as VAT data and ndustral producton ndces, quarterly gross value added s estmated by ndustry for more than 30 branches (the detaled methodology to estmate quarterly value added by ndustry s descrbed n the specfc Manual mentoned under 2.1 above). Ths data set s translated nto a sector breakdown usng approprate keys. and domestc sub sectors (S.11, S.12, S.13, S.1M) The keys that allow dstrbutng the quarterly value added amongst the nsttutonal sectors are establshed usng, for each ndustry, the annual breakdown of the value added between the nsttutonal sectors. Ths means that quarterly value added s allocated among the sectors S.11, S.12, S.13 and S.1M accordng to the allocaton of the corpondng amount regstered on an annual bass. VA S b VA S b,

15 15. wth S b, VA VA VA S b b x VA b, where S : nsttutonal sector b : actvty branch : quarter The followng keys were d, for each quarter of 2012, to dstrbute the quarterly value added by ndustry amongst the nsttutonal sectors: Table 2 - Breakdown of value added by ndustry between nsttutonal sectors (annual keys for 2012) S.11 S.12 S.13 S.14(*) S.15(*) Agrculture, fotry and fshng 33.1% 66.9% Mnng and quarryng 99.7% 0.3% Food products, beverages and tobacco 95.9% 4.1% Textle and leather 98.2% 1.8% Wood, paper product and prntng 97.7% 2.3% Coke, refned petroleum products 100% Chemcals 100% Pharmaceutcal products 100% Rubber, plastc products and other nonmetallc products Basc metals and fabrcated metal products 99.6% 0.4% 98.6% 1.4% Computer electronc and optcal products 99.8% 0.2% Electrcal equpment 99.8% 0.2% Other machnery and equpment 99.8% 0.2% Transport equpment 99.9% 0.1% Furnture 94.3% 5.7% Repar 99.0% 1,0% Electrcty and gas 100% Water supply, sewage and waste management 82.2% 17.3% 0.4% Constructon 88.5% 11.5% Wholesale and retal trade 93.8% 6.2% Transportaton and storage 72.6% 27.0% 1.0% Accommodaton and food actvtes 76.8% 23.2% Publshng audo-vsual and broadcastng actvtes 81.5% 17.7% 0.9% Telecommuncaton 98.8% 0.2% Computer and nformaton servce actvtes Fnancal ntermedaton 100% Insurance 100% 98.2% 1.8% Fnancal auxla 96.4% 3.6% Real estate actvtes 97.7% 2.3% Dwellng servces 100% Legal, accountng, consultancy, archtecture and engneerng actvtes 53.3% 0.1% 46.6% Scentfc earch and development 88.7% 0.2% 11.1% Advertsng, other professonal scentfc and techncal actvtes, veternary 83.9% 16.1% Admnstratve and support servces 95.9% 4.2%

16 16. Publc admnstraton 100% Educaton Human health Socal work Art, entertanment and recreaton Other servce actvtes Domestc servces 100% (*) The sectors S.14 and S.15 are merged n QSA but the calculaton of quarterly value added s frst made separately. When annual data for value added by nsttutonal sector are not yet avalable, the keys calculated accordng to the last avalable year are kept unchanged 10. Ths can lead, for sector S.13, to an estmate of the value added whch s dfferent from the fgure avalable n the government data. As these data have prorty and, n order to avod dsturbng the GDP (producton sde), the dfference n the valuaton of B.1g-S13 s counterbalanced by an equvalent correcton (wth the opposte sgn) n the valuaton of the value added for non-fnancal corporatons (B.1g-S11). Rest of the world Fnal consumpton expendture (P.3) Quarterly pre-exstng data are d (man aggregates and general government accounts). A revew of the detaled methods to establsh these data can be found n the above mentoned documents (cf. 2.1 above). Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) Data are drectly taken from the quarterly general government accounts set up at t+85 days. General government accounts nclude P.3 breakdown between P.31 and P.32 (ndvdual and collectve consumpton expendture). Hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly man aggregates data set released 60 days after the end of the quarter (part of Table 01 of the transmsson program). Data for S.1 are calculated as the sum of domestc sectors. As the data d for general government (cf. supra) are updated compared to the one d to comple the man aggregates, fnal consumpton expendture for the total economy s not necessarly n lne for the latest quarters wth the one appearng n the last released man aggregates Ths means that the same dstrbuton keys are appled for a maxmum of 9 successve quarters: when annual accounts are publshed at the end of September of year Y, the keys calculated for the last publshed year Y-1 are d to splt up the value added for the four quarters of Y-1, the four quarters of Y and the frst quarter of Y+1. When QSA for the second quarter of Y+1 wll be released, newly released annual keys relatng to year Y wll be d. 11 When necessary, a counterpart adjustment s brought to changes n nvento (part of P.5N); cf

17 Gross captal formaton (P.5) Gross captal formaton s defned as the sum of gross fxed captal formaton (P.5) and the changes n nvento together wth acqustons less dsposals of valuables (P.5N) Gross fxed captal formaton (P.51) Quarterly pre-exstng data are d (man aggregates and general government accounts). A revew of the detaled methods to establsh these data can be found n the above mentoned documents (cf. 2.1 above). The avalable quarterly data n the man aggregates are: housng nvestments; government nvestments ; 'prvate' nvestments (excl. housng). Housng nvestments have to be totally mputed to hoholds. QSA requre breakng down 'prvate' nvestments (excl. housng) between non-fnancal corporatons (S.11), fnancal corporatons (S.12) and hoholds ncl. NPISH's (S.1M). Publc nvestments are taken from general government accounts. Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M), t of the world (S.2) Quarterly 'prvate' nvestments (excl. housng) are broken down between non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds (S.14) and NPISH's (S.15) by consderng the annual share of each of these sectors n the corpondng annual aggregate. Those data for the sector S14 refer to nvestments attrbuted to self-employed workers. They are added to housng nvestments - avalable n the man aggregates and whch are totally attrbuted to hoholds - to estmate the total nvestments for the sector S.14. The splttng of quarterly prvate nvestment between nsttutonal sectors s not proceeded n a detaled approach by ndust as quarterly fgu of nvestment by ndust are not avalable. The quarterly splttng of prvate nvestment between sectors s made usng the followng formula: S I S I I x I wth I S I referrng to the annual share of the nsttutonal sector S n the total annual prvate nvestment; where I : prvate nvestment (excl. housng) S : nsttutonal sector (S11, S12, S14 or S15) : quarter Data are drectly taken from the quarterly general government accounts set up at t+85 days,

18 18. Data for S.1 are calculated as the sum of domestc sectors. As the data d for general government (cf. supra) are updated compared to the one d to comple the man aggregates, gross fxed captal formaton for the total economy s not necessary n lne for the latest quarters wth the one appearng n the last released man aggregates Changes n nvento and acqustons less dsposals of valuables (P.5N) Quarterly pre-exstng data, comng from the man aggregates, are d as startng pont for the whole economy. For the latest quarters, dscrepances appear between the fgu comng from the quarterly man aggregates (Table 1) and those comng from the general government accounts. Dscrepances concern government fnal consumpton (P.3_S13), government nvestments (P.51_S.13) and taxes less subsdes on products (D.21-D31_S.1). Government data s always consdered as a prorty. Hence, an adjustment s brought n the total changes n nvento, n order to offset the dscrepancy on the expendture sde (for P.3_S13 and P.51_S.13), so that the GDP n QSA s consstent wth the fgu publshed n the man aggregates, except for updated taxes less subsdes on products. As a consequence, quarterly GDP data n QSA are closed to the one transmtted n Table 1 of the transmsson program but ts decomposton nto the dfferent catego of expendture can be dfferent for the latest quarters 13. Ths adjustment and remanng dscrepancy have a lmted sze. Quarterly data for the whole economy have to be broken down amongst the nsttutonal sectors S.11, S.12, S.13 and S.1M. Non-fnancal corporatons (S.11) Ths s the dual tem for transacton P.5N. For each quarter : P.5N _ S.11 P.5N _ S.1 P5N_ S.12 P.5N _ S13 P.5N _ S.1M Fnancal corporatons (S.12) The small corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. As specfed n , when annual accounts are not yet avalable, quarterly data for transactons estmated by a smoothng method are bult up by extendng the lnear trend. Data are drectly taken from the quarterly general government accounts set up at t+85 days. Hoholds and NPISH's (S.1M) The small corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. 12 When necessary, a counterpart adjustment s brought to changes n nvento (part of P.5N); cf In the case of revsed annual (hence quarterly) general government data that are not ncorporated yet n the man aggregates, more extended dscrepances (for the whole perod of transmsson) can occur. In ths case, adjustments to changes n nvento should be brought for each quarter of the perod of transmsson.

19 19. As specfed n , when annual accounts are not yet avalable, quarterly data for transactons estmated by a smoothng method are bult up by extendng the lnear trend. For each quarter : P.5N _ S.1 P.5N _ S.1 QSA Table1 P.51_ S.13Table1 P.51_ S.13Table25 D.21 D.31 _ S.1 D.21 D.31 Rest of the world Table25 P.3 _ S.13 _ S.1 Table1 Table1 P.3 _ S.13 Table Exports (P.6) and mports (P.7) of goods and servces Data for P.6 and P.7, but also for P.61 (exports of goods), P.62 (export of servces), P.71 (mports of goods) and P.72 (mports of servces), are drectly taken from the quarterly man aggregates (part of Table 01 of the transmsson program). A revew of the detaled methods to establsh these data can be found n the above mentoned document (cf. 2.1 above) Compensaton of employees (D.1) (A) Uses sde and domestc sub sectors (S.11, S.12, S.13, S.1M) Quarterly pre-exstng data, comng from the man aggregates publshed 60 days after the end of the quarter, are d as startng pont for the whole economy. Compensaton of employees pad by the government s taken from the general government accounts. The dfference between those two aggregates repents the wage bll of the prvate sector (S.1-S.13) that has to be splt up amongst the nsttutonal sectors S.11, S.12 and S.1M. Compensaton of employees pad by the general government s estmated wthn the general government accounts. The dfference between the total wage bll and ths compensaton has to be splt up amongst the non-government nsttutonal sectors. Ths breakdown s made by usng an ndcator bult up on quarterly nformaton comng from the NSSO. Ths nformaton s avalable, n a frst verson, wth a tme lag of 4 months and, n a second verson, wth a tme lag of 7 months 14. Dependng on the qualty and the exhaustveness of the NSSO ndcator, the annual total for each nsttutonal sector arsng from ths method dverges more or less from the annual data that prevals n the annual ESA2010 accounts. The quarterly ults have to be adjusted n a second step to ft the annual released data. For each 'prvate' nsttutonal sector, the dfference between the sum of the four quarters calculated accordng to the NSSO ndcator and the annual released data s dstrbuted accordng to the weght of the quarter wthn the year. 14 Quarterly data for the whole economy (S.1) publshed n the man aggregates are based on the same nformaton comng from the NSSO.

20 20. The very lmted dual dfference that could reman between the sum of the three sectors (S.11, S.12 and S.1M) and the total constrant for S.1-S.13 s attrbuted to wages pad by non-fnancal corporatons (S.11). When annual data by nsttutonal sector are not yet avalable, quarterly estmates are bult up by consderng the evoluton of the NSSO ndcator compared to the corpondng quarter of the prevous year and adjustng the ult to ft the quarterly released data for the whole domestc economy (man aggregates). For the most recent quarter, for whch no nformaton comng from the NSSO s avalable, the ndcator s prolonged on the bass of the development of the wage bll estmated for the whole economy and the pectve developments by nsttutonal sector observed n the recent past. Table 3 - Breakdown of the quarterly domestc compensaton of employees pad between nsttutonal sectors (estmates for 2012) S.11 S.12 S.13 S.1M Q ,0% 5,3% 25,0% 2,6% Q ,7% 5,4% 24,2% 2,8% Q ,0% 4,9% 25,3% 2,8% Q ,7% 4,8% 23,7% 2,7% The data are calculated on the bass of Bop data. When annual data are avalable, the quarterly profle s defned by usng, as ndcator, nformaton comng from the Bop. When annual data are not yet avalable, the data comng from Bop s drectly d. For recent years, dscrepances between NA data and Bop data for compensaton of employees are neglgble. (B) Resources sde Non-fnancal corporatons (S.11), fnancal-corporatons (S.12), general government (S.13) Hoholds and NPISH's (S.1M) The wage bll for the whole domestc economy [cf. supra (A)] adjusted for compensaton pad and receved by the t of the world totally benefts to the sector of hoholds. For each quarter : D.1_ S1M D.1_ S.1 D.1_ S.2 D.1_ S.2 Sum of domestc sectors (here lmted to S.1M). The data are calculated on the bass of Bop data. When annual data are avalable, the quarterly profle s defned by usng, as ndcator, nformaton comng from the Bop. When annual data are not yet avalable, the data comng from Bop s drectly d. For recent years, dscrepances between NA data and Bop data for compensaton of employees are neglgble.

21 Taxes on producton and mports (D.2) Taxes on producton and mports are defned as the sum of taxes on products (D.21) and other taxes on producton (D.29) Taxes on products (D.21) (A) Uses sde Quarterly taxes on products that are estmated on the ources sde for the total economy (S.1) and the t of the world (S.2) [cf. nfra (B)] are entrely pad by the total economy (S.1) wthout dstncton of nsttutonal sectors, n accordance wth ESA2010. For each quarter : D.21_ S1 D.21_ S.1 D.21_ S.2 (B) Resources sde Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts set up at t+85 days. Sum of domestc sectors (here lmted to S.13). As the data d for general government (cf. supra) are updated compared to the one d to comple the man aggregates, taxes on products for the total economy are not necessary n lne for the latest quarters wth the one appearng n the last released man aggregates 15. Data are drectly taken from the quarterly general government accounts set up at t+85 days. Ths data set reports taxes that are pad to EU nsttutons (manly part of taxes from the common agrcultural polcy and customs dutes) Other taxes on producton (D.29) (A) Uses sde 15 If requred, a counterpart adjustment s brought to changes n nvento (part of P.5N); cf

22 22. Source No nformaton on the decomposton of other taxes on producton pad by the dfferent nsttutonal sectors s avalable on a quarterly bass. Estmates must be made. Non-fnancal corporatons (S.11) Ths s the dual tem for transacton D.29. For each quarter : D.29_ S.11 D.29_ S.1 D.29_ S.12 D.29_ S13 D.29_ S.1M Fnancal corporatons (S.12) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Actually, the transacton s nl n Belgan accounts. Hoholds and NPISH's (S.1M) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. The annual amounts for other taxes on producton pad by hoholds are qute large and manly concern taxes on the ownershp or of buldngs. The smoothng treatment to estmate quarterly data can be consdered as approprate accordng to the accrual recordng concept, as those taxes refer to a producton of housng servces that are evenly spread over tme. Other taxes on producton pad by the total economy are equal to the amount receved by the general government. (B) Resources sde Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M), t of the world (S.2) Data are drectly taken from the quarterly general government accounts set up at t+85 days.

23 23. Sum of domestc sectors (here lmted to S.13) Subsdes (D.3) Subsdes are defned as the sum of subsdes on products (D.31) and other subsdes on producton (D.39) Subsdes on products (D.31) (A) Uses sde Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts set up at t+85 days. Sum of domestc sectors (here lmted to S.13). As the data d for general government (cf. supra) are updated compared to the one d to comple the man aggregates, subsdes on products for the total economy are not necessarly n lne for the latest quarters wth the one appearng n the last released man aggregates 16. Data are drectly taken from the quarterly general government accounts set up at t+85 days. These report subsdes that are receved from EU nsttutons. (B) Resources sde Quarterly subsdes on products that are estmated on the s sde for the total economy (S.1) and the t of the world (S.2) [cf. supra (A)] are entrely receved by the total economy (S.1) wthout dstncton of nsttutonal sectors, n accordance wth ESA2010. For each quarter : D.31_ S1 D.31_ S.1 D.31_ S Other subsdes on producton (D.39) (A) Uses sde 16 If requred, a counterpart adjustment s brought to changes n nvento (part of P.5N); cf

24 24. Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts set up at t+85days. Sum of domestc sectors (here lmted to S.13). Data are drectly taken from the quarterly general government accounts set up at t+85 days. (B) Resources sde Source No nformaton on the decomposton of other subsdes on producton receved by the dfferent nsttutonal sectors s avalable on a quarterly bass. Estmates must be made. Non-fnancal corporatons (S.11) Ths s the dual tem for transacton D.39. For each quarter : D.39_ S.11 D.39_ S.1 D.39_ S.12 D.39_ S13 D.39_ S.1M Fnancal corporatons (S.12) The lmted corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts set up at t+85 days. Hoholds and NPISH's (S.1M) The lmted corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Other subsdes on producton receved by the total economy are equal to the amount pad by the general government and the t on the world. For each quarter :

25 25. D.39_ S.1 D.39_ S.13 D.39_ S Intet (D.41) Unless otherwse stated, quarterly ntet flows are defned after FISIM allocaton. The only pre-exstng quarterly nformaton comes from the quarterly general government accounts: ntet pad and receved by the sector S.13 must be consdered as constrants n QSA. For the other nsttutonal sectors, as well as for the whole economy, quarterly estmates of ntet flows have to be produced n pect of the dentty accordng to whch the sum of the quarters must equal the annual data. For the years from 2003 onwards, ths dentty s met as the released annual data are based on the summaton of quarterly estmates for ntet receved and pad by each nsttutonal sector (ntet matrx on quarterly bass). For the quarters of the years before 2003, quarterly matrces have not been bult up beca ths work proved to be very tme-consumng. A smplfed ad hoc procedure, whch reles on smoothng trends, s appled to get quarterly estmates of total ntet receved and pad by each nsttutonal sector wthout nformaton on the counterpart sector. Quarterly estmates before 2003 (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts (these data are "consoldated".e. they exclude ntet flows wthn enttes of the government). Hoholds and NPISH's (S.1M) Ths s the dual tem for transacton D.41, on the s sde. For each quarter : D.41_ S.1M D.41_ S.1 D.41_ S.11 D.41_ S12 D.41_ S.13 The corpondng annual fgure (excludng ntet payments wthn enttes of the general government sector) s smoothed out over the four quarters of the year, by applyng a lnear trend. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend.

26 26. Data are drectly taken from the quarterly general government accounts (these data are "consoldated".e. they exclude ntet flows wthn enttes of the government). Hoholds and NPISH's (S.1M) Ths s the dual tem for transacton D.41 on the ources sde. For each quarter : D.41_ S.1M D.41_ S.1 D.41_ S.11 D.41_ S12 D.41_ S.13 Quarterly ntets receved by the total economy are defned accordngly to quarterly ntets pad by the domestc sectors adjusted for net ntet wth the t of the world. For each quarter : D.41_ S.1 D.41_ S.1 D.41_ S.2 D.41_ S.2 Quarterly estmates for Avalablty of annual accounts As annual data relatng to ntet flows are bult up on the bass of quarterly matrces combnng the s sde and the ources sde, these quarterly matrces are d for the purpose of QSA and are fully consstent wth the corpondng annual fgu. Quarterly matrces are defned before FISIM allocaton and take nto account the followng constrants: quarterly ntet pad and receved by the central bank, other monetary fnancal nsttutons and mutual funds. The quarterly avalable nformaton, comng from accountng data and reportng schemes, does not nclude nformaton about counterpart sectors (except for the nterbank payments that are taken on board); quarterly ntet pad and receved by the general government wthout nformaton about counterpart sectors; annual ntet flows for nsurance corporatons and penson funds that are assessed manly on the bass of nformaton comng from the Fnancal supervsory and market authorty (FSMA) wthout nformaton about counterpart sectors. The breakdown of quarterly ntet flows by counterpart sector s estmated on the bass of the outstandng amounts of assets and labltes (nstruments F.2, F.3 and F.4) and assocated ntet rates. The most detaled level of fnancal accounts classfcaton s d so that the most approprate return by nstrument can be appled. Adjustments to ft the above mentoned constrants are made to the estmated matrx by combnng a manual balancng procedure and, for small dual amounts, a RAS procedure 17. As the quarterly matrces relate to ntet flows before FISIM allocaton, correctons have to be made n order to reclassfy part of the ntet payments as payments for fnancal servces. They are estmated on the bass of quarterly detaled data about assets and labltes comng from the fnancal accounts and assocated ntet rate margns (FISIM calculaton at quarterly level). For tme-span , the ntet matrces and the FISIM calculatons are establshed on the bass of ESA1995 sectorzaton and concepts. It was not possble to re-buld all matrces back to 2003 on the bass of ESA2010. It was decded to keep ESA1995 matrces, and to adjust for new sector classfcaton and revsons n sectors S.13 and S.2, knowng that there were no major methodologcal change between ESA1995 and ESA2010 n the feld of ntet or FISIM. 17 In contrast to procedu appled for many transactons, no specfc dual sector s defned to balance the system.

27 27. Quarterly estmates from 2013 on From 2013 on, quarterly ntet matrces on the bass of ESA2010 sector classfcaton have been set up. The buldng prncples are the same as prevously: for each sector, ntet flows by counterpart sector s estmated on the bass of the outstandng amounts of assets and labltes (nstruments F.2, F.3, F.4 and F.8 recorded n fnancal accounts) multpled by an ad hoc assocated ntet rates. For sectors S.13, S.121, S.122, S.123, S.124, S.128 and S.129, total ntets are known from other sources and taken nto account. FISIM calculaton s estmated accordng to the same prncples and data sources. Annual ntet flows are calculated as the sum of quarterly ntet flows, ensurng perfect consstency between ASA and QSA for ntet Property ncome other than ntet (D.4N) Property ncome other than ntet s the summng up of the components: D.42, D.43, D.441/2, D.443 and D Dvdends (D.421) When annual data are avalable, quarterly profles of dvdends are defned usng nformaton comng from the balance sheets of Belgan companes and from the Bop. For the latest quarters (when annual accounts are not yet avalable), an ad hoc methodology has been developed. Avalablty of annual accounts A quarterly allocaton key s calculated for dvdends pad and dvdends receved. (A) Uses Non-fnancal corporatons (S.11), fnancal corporatons (S.12) The quarterly profle s determned on the bass of nformaton comng from the balance sheets of companes and Bop. Accordng to the collected nformaton, the major part of dvdends s pad by Belgan companes durng the second quarter of the year. Table 4 - Quarterly profle of dvdends pad by Belgan corporatons Non-fnancal corporatons (S.11) Q1 8% 8% 8% Q2 63% 63% 63% Q3 16% 16% 16% Q4 14% 14% 14% Fnancal corporatons (S.12) Q1 12% 13% 14% Q2 61% 60% 60% Q3 14% 14% 13% Q4 13% 13% 13%

28 28. Box - Addtonal explanaton for the estmaton of quarterly profle of dvdends pad by dent enterprses Non-fnancal corporatons (S11) The ndvdual annual accounts of corporatons contan nformaton on the date of the annual general meetng of shareholders (AGM). As a rule, dvdends are not pad out untl several weeks after the AGM, but precse nformaton on ths s not avalable. On the bass of the date of the AGM, the quarterly pattern of pad dvdends s calculated separately for the full-format and short-format accounts. The bulk of the dvdends s pad out n the second quarter. The dvdend pattern for the short format accounts s more or less the same as that for the full format. Dvdends pad out n year t relate to actvtes n t-1. The allocaton key s therefore lagged by 1 year. Central Bank (S.121) Informaton s avalable on the annual account of the Natonal Bank. Depost takng corporatons (S.122) The quarterly allocaton key for S.122 s based on the date of fllng the annual accounts. Snce no annual accounts are flled n June, there s no need for any shft from the second to the thrd quarter, n contrast to S.11. The allocaton key appears to be qute constant over tme. Other fnancal sub-sectors (S123, S124, S125, S126, S127) The calculatons are performed n the same way as for S.11., hoholds and NPISH's (S.1M) Sum of domestc sectors An agreement was reached between natonal accounts experts and Bop experts to dentcal quarterly allocaton keys for dvdends pad by the t of the world. The general profle that has been set, accordng to a combnaton of companes' balance sheets and Bop nformaton, ensu a global consstency of dvdend flows amongst all the nsttutonal sectors (ncludng plausble ults for the dual sector S.1M on the ources sde). These common keys apply from 2004 onwards. The agreed general pattern for quarterly profles has also been adopted n QSA for the years before 2004, so that constant allocaton keys are d for Table 5 - Quarterly profle of dvdends pad by the t of the world accordng to natonal accounts Q1 10% 10% 10% Q2 64% 64% 64% Q3 13% 13% 13% Q4 13% 13% 13% (B) Resources

29 29. and domestc sub sectors (S.11, S.12, S.13, S.1M) Quarterly dvdends receved by the total economy are defned accordngly to quarterly dvdends pad by the domestc sectors adjusted for net dvdend flows wth the t of the world. For each quarter : D.421_ S.1 D.421_ S.1 D.421_ S.2 D.421_ S.2 Quarterly dvdends receved by the total economy are allocated to the domestc sectors as follows: dvdends receved by the government (S13) are drectly taken from the quarterly general government accounts set up at t+85 days; D. 421_ S.1 D.421_ S. 13 s allocated pro rata among the sectors the dual amount S.11, S.12 and S.1M accordng to the amounts receved on an annual bass. As for dvdends pad by the t of the world [cf. supra (A)], quarterly allocaton keys are common to QSA and Bop from 2004 onwards. For the former perod, quarterly dvdends receved by the t of the world are calculated n QSA wth the ultmate objectve to keep net dvdend flows wth the t of the world (arsng from quarterly Bop data) unchanged. The quarterly profle of dvdend receved by the t of the world n QSA s the followng: Table 6 - Quarterly profle of dvdends receved by the t of the world accordng to natonal accounts Q1 12% 10% 10% Q2 57% 58% 58% Q3 17% 17% 17% Q4 15% 15% 15% Unavalablty of annual accounts If the dvdends calculated for the purpose of the annual natonal accounts are not yet avalable, an estmate s produced. (A) Uses Non-fnancal corporatons (S.11) The method reles on the fact that dvdends pad n year Y by dent corporatons depend on the proft generated n Y-1. The lnk that can be establshed between those two aggregates has been analysed on the bass of the followng regson 18 : D.421 a b B.2g. Ths regson has been examned on an annual bass (publshed data from 1995 onwards) for sx ndustry groups. When t proved not to be sgnfcant, a movng average of the rato D.421/B.2g s appled. The annual projecton of dvdends s, n a second step, allocated amongst the quarters on the bass of the quarterly pattern of the latest avalable annual data. Y 1 Y 18 Ths regson s a smplfed vew of the tems underlyng payment of dvdends.

30 30. Table 7 - Method d to estmate annual dvdends (when annual natonal accounts are not yet avalable) NACE 01_05 agrculture/fshe D.421 (Y) = D.421 (Y-1) 14_37 manufacturng regson on B.2g (Y-1) 40_41 electrcty, gas, water movng average of D.421/B.2g over 3 years 45 constructon movng average of D.421/B.2g over 3 years 50_52 trade regson on B.2g (Y-1) 55_90 servces regson on B.2g (Y-1) Fnancal nsttutons (S.12) Snce fnancal nsttutons have to meet strct and tmely reportng oblgatons, suffcent nformaton s generally avalable to estmate dvdends pad durng the latest quarters. Dvdends relatng to quarters of year Y+1 after the fnal year of avalable natonal accounts are estmated on the bass of accountng reports for year Y that are already avalable around Aprl Y+1., hoholds and NPISH's (S.1M) Sum of domestc sectors. Dvdends pad by the t of the world (portfolo and drect nvestment) to domestc sectors are estmated on the bass of data comng from the Bop. (A) Resources and domestc sub sectors (S.11, S.12, S.13, S.1M) Dvdends receved by Belgan dents are calculated as a dual fgure per quarter, n the same way as for the years for whch annual data are already avalable (cf. supra). They are allocated to sectors va the outstandng amounts of sha and other equty (nformaton avalable n the fnancal accounts). Both dvdends ultng from portfolo and drect nvestment are estmated on the bass of data comng from the Bop Wthdrawals from the ncome of quas-corporatons (D.422) Quarterly amounts of wthdrawals from the ncome of quas-corporatons are drectly taken from the quarterly general government accounts set up at t85 days (S13/ources sde) wth, as full and sngle counterpart, expendture for dent non-fnancal corporatons (S.11/s sde) Renvested earnngs on drect foregn nvestment (D.43) When annual data are avalable, quarterly profles of renvested earnngs on drect foregn nvestment are defned usng nformaton comng from the balance sheets of Belgan companes and from the Bop. For the latest quarters (when annual accounts are not yet avalable), an ad hoc methodology s appled.

31 31. Avalablty of annual accounts A quarterly allocaton key s calculated for renvested earnngs pad and renvested earnngs receved. (A) Uses Non-fnancal corporatons (S.11) The calculatons nclude the followng steps: (a) For each year snce 1999 the percentage allocaton per quarter of the value added per ndustry (ndustry breakdown of the quarterly accounts) s calculated. (b) The followng data from the drect nvestment enterprses ncluded n the database d to calculate the annual fgu for D.43 are aggregated at NACE 2-dgt level: the net current ult (current operatng proft) x control percentage dvdends x control percentage (c) The corpondng quarterly allocaton key (cf. (a)) s appled to the annual fgure of the "net ult x control%" under (b), whch gves the quarterly net ult of the drect nvestment enterprses per NACE 2-dgt headng. (d) The quarterly pattern of dvdends pad per NACE 2-dgt headng s avalable from the nformaton n the annual accounts. (e) The corpondng quarterly allocaton key s appled to the annual fgure of the "dvdends x control %" under (b), gvng the dvdends per quarter. (f) Per NACE 2-dgt headng, ths gves per quarter: D.43 = net ult dvdends. Summng across all sectors gves 4 quarterly amounts whch, except for mnor roundng errors and any adjustment effects, corpond to the fgure calculated annually for D.43, and can therefore be appled as allocaton key to that annual fgure. The assumptons underlyng these calculatons are: the allocaton of the gross operatng surplus (for whch the value added B.1g s d as a proxy) for all enterprses (cf. (a)) s repentatve for drect nvestment enterprses. The gross operatng surplus s tself a proxy for the net current ult. the nformaton on the quarterly pattern of dvdends pad, based on nformaton for all annual accounts avalable at the Central Balance Sheet Offce, s repentatve for drect nvestment enterprses. Fnancal corporatons (S.12) In the case of fnancal corporatons (S.12), the quarterly patterns of dvdends pad, calculated for the purpose of the quarterly accounts, are taken as a proxy for the quarterly patterns of the drect nvestment enterprses. The quarterly patterns of the net ult for S.122, S.125, S126 and S.127 are calculated n a smlar way as for S.11,.e. on the bass of the quarterly pattern of the value added of NACE (fnancal nsttutons + fnancal auxla). For the quarterly pattern of the net ult for S.128, the quarterly pattern of NACE 65 (nsurance sector) s d.

32 32. For each sub sector of S.12, the quarterly fgu for D.43 thus obtaned (= net ult - dvdends) are d as the allocaton key for the purpose of the quarterly allocaton of the D.43 calculated n the annual natonal accounts., hoholds and NPISH's (S.1M) Sum of domestc sectors. Rest of the world s the counterpart sector of corpondng quarterly transactons recorded as ources for Belgan corporatons. For each quarter : D.43_ S2 D.43_ S.11 D.43_ S.12 (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12) The quarterly Bop data on the foregn drect nvestment survey for non-fnancal corporatons and fnancal nsttutons are taken on board., hoholds and NPISH's (S.1M) Sum of domestc sectors. Rest of the world s the counterpart sector of corpondng quarterly transactons recorded as s for Belgan corporatons. For each quarter : D.43_ S2 D.43_ S.11 D.43_ S.12 Unavalablty of annual accounts (A) Uses Non-fnancal corporatons (S.11), fnancal corporatons (S.12) Currently, quarterly estmates relatng to the latest year for whch annual accounts are avalable are kept unchanged for the corpondng quarters relatng to the followng years., hoholds and NPISH's (S.1M) Sum of domestc sectors.

33 33. Rest of the world s the counterpart sector of corpondng quarterly transactons recorded as ources for Belgan corporatons. For each quarter : D.43_ S2 D.43_ S.11 D.43_ S.12 (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12) Currently, quarterly estmates relatng to the latest year for whch annual accounts are avalable are kept unchanged for the corpondng quarters relatng to the followng years., hoholds and NPISH's (S.1M) Sum of domestc sectors. Rest of the world s the counterpart sector of corpondng quarterly transactons recorded as s for Belgan corporatons. For each quarter : D.43_ S2 D.43_ S.11 D.43_ S Property ncome attrbutable to nsurance polcy holders (D.441) +Investment ncome payable on penson enttlements (D.442) No quarterly nformaton s avalable and a smple smoothng method s d. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts set up at t+85 days. Hoholds and NPISH's (S.1M) Sum of domestc sectors. (B) Resources sde

34 34. Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Quarterly general government accounts set up a t+85 days, gve data on property ncome attrbuted to nsurance polcy holders (D.441) together wth rents (D.45). For the purposes of QSA, ths nformaton s allocated pro rata amongst both transactons accordng to the amounts regstered on an annual bass. When annual accounts are not yet avalable, the most recent annual dstrbuton key s kept unchanged. Hoholds and NPISH's (S.1M) Ths s the dual tem for transacton D.441+D442. For each quarter : D. 441/ 2_ S.1M D.441/ 2_ S.1 D.441/ 2_ S.11 D.441/ 2_ S12 D.441/ 2_ S. 13 Quarterly property ncome attrbuted to nsurance polcy holders receved by the total economy are defned accordng to quarterly property ncome attrbuted to nsurance polcy holders pad by the domestc sectors adjusted for net corpondng flows wth the t of the world. For each quarter : D. 441/ 2_ S.1 D.441/ 2_ S.1 D.441/ 2_ S.2 D.441/ 2_ S Investment ncome attrbutable to collectve nvestment fund (CIF) shareholders (D.443) From 2009Q1 onwards, the estmaton of nvestment ncome attrbutable to CIF shareholders s calculated on quarterly bass. The annual data are calculated as the sum of quarterly estmates. As regards dent CIF, the nvestment ncome s collected by the supervsory authortes on quarterly bass. The ncome s allocated among counterpart sectors proportonally to the holdng of dent CIF sha of each sector as recorded n the quarterly fnancal accounts. As regards non-dent CIF, the quarterly nvestment ncome comes from the Bop data. Ths total ncome s allocated among counterpart sectors proportonally to the holdng of non-dent CIF sha of each sector as recorded n the quarterly fnancal accounts. The estmaton s very detaled, as t dstngushes monetary/non-monetary CIF, D4431/D4432 and each sub-sector of S.12. Before 2009, t was not possble to mplement ths detaled calculaton method, as Bop-BPM6 data are not avalable and as the estmaton s establshed on annual bass. Consequently, a smplfed methodology s d,.e. the annual fgure for each sector (ource or ) s spread over quarters usng a key. Ths key s equal to 2/3 of the corpondng key for ntet and 1/3 of the corpondng key for dvdends. Ths calculaton s made for each se (D.443 pad by S12, D.443 pad by S.2 and D.443 receved by each sector). A fnal adjustment s appled to ensure that total ources are equal to total s.

35 35. (A) Uses sde Fnancal corporatons (S.12), t of the world (S.2) The data comes drectly from sources as explaned above. Non-fnancal corporatons (S.11), general government (S.13), hoholds and NPISH's (S.1M) Equal to S.12. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M), t of the world (S.2) The data comes drectly from sources as explaned above. The transacton s nl n Belgan accounts. Quarterly property ncome attrbutable to CIF shareholders receved by the total economy are defned accordng to quarterly property ncome attrbuted to CIF shareholders pad by the domestc sectors adjusted for net corpondng flows wth the t of the world. For each quarter : D. 443_ S.1 D.443_ S.1 D.443_ S.2 D.443_ S Rents (D.45) No quarterly nformaton s avalable and a smple smoothng method s d. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Nl n Belgan accounts.

36 36. Sum of domestc sectors. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Quarterly general government accounts set up at t+85 days, gve data on property ncome attrbuted to nsurance polcy holders (D.44) together wth rents (D.45). For the purposes of QSA, ths nformaton s allocated pro rata amongst both transactons accordng to the amounts regstered on an annual bass. When annual accounts are not yet avalable, the most recent annual dstrbuton key s kept unchanged. Hoholds and NPISH's (S.1M) Ths s the dual tem for transacton D.45. For each quarter : D.45_ S.1M D.45_ S.1 D.45_ S.11 D.45_ S.12 D.45_ S13 Quarterly rents receved by the total economy are defned accordng to quarterly rents pad by the domestc sectors, as transactons wth the t of the world are non-exstent. For each quarter : D.45_ S.1 D.45_ S Current taxes on ncome, wealth, etc. (D.5) Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12) Quarterly general government accounts gve nformaton about current taxes pad by corporatons, whether these are non-fnancal or fnancal enterprses. For the purposes of QSA, ths nformaton s allocated pro rata amongst both sub sectors accordng to the amounts pad on an annual bass. When annual accounts are not yet avalable, the most recent annual dstrbuton key s kept unchanged.

37 37. Quarterly general government accounts gve nformaton about current taxes wthn the government. Hoholds and NPISH's (S.1M) Quarterly general government accounts gve nformaton about current taxes pad by hoholds and NPISH's. Sum of domestc sectors. Quarterly general government accounts gve nformaton about current taxes pad by non-dents. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts. Sum of domestc sectors (here lmted to S.13). Current taxes pad by dents to the t of the world are taken from the quarterly Bop of the European Insttutons, publshed on Eurostat webste Socal contrbutons (D.61) Actual socal contrbutons (D61x = D D D D.61SC) The actual socal contrbutons relatng to publc schemes come from quarterly general government accounts. For prvate schemes, no nformaton s at pent avalable on a quarterly bass and a smple smoothng method of annual data s d. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), general government (S.13) Hoholds and NPISH's (S.1M) Total actual socal contrbutons on the ources sde of the whole domestc economy [cf. nfra (B)] adjusted for net corpondng flows wth the t of the world are totally supported by hoholds and NPISH's. For each quarter :

38 D. 61_ S1M D.61_ S.1 D.61_ S.2 D.61_ S Sum of domestc sectors (here lmted to S.1M). The data are calculated on the bass of Bop data. When annual data are avalable, the quarterly profle s defned by usng, as ndcator, nformaton comng from the Bop. When annual data are not yet avalable, the data comng from Bop s drectly d. Due to harmonzaton procedure between natonal accounts and Bop, the dscrepances between both statstcs for ths transacton are neglgble for the more recent years. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts. Hoholds and NPISH's (S.1M) Sum of domestc sectors. The data are calculated on the bass of Bop data. When annual data are avalable, the quarterly profle s defned by usng, as ndcator, nformaton comng from the Bop. When annual data are not yet avalable, the data comng from Bop s drectly d. Due to harmonzaton procedure between natonal accounts and Bop, the dscrepances between both statstcs for ths transacton are neglgble for the more recent years Imputed socal contrbutons (D.612) The quarterly nformaton comes from quarterly general government accounts. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), general government (S.13), t of the world (S.2) Hoholds and NPISH's (S.1M)

39 39. Imputed socal contrbutons on the ources sde of the whole domestc economy [cf. nfra (B)] adjusted for mputed socal contrbutons pad to non-dents are totally supported by hoholds and NPISH's. For each quarter : D.612_ S1M D.612_ S.1 D.612_ S.2 Sum of domestc sectors (here lmted to S.1M). (B) Resources sde Non-fnancal corporatons (S.11) The followng accountng dentty s d: For each quarter : D. 612_ S11 D.62c_ S11 D.611_ S11 ( partm ) (D62C = socal beneft drectly pad by the employer) Fnancal corporatons (S.12), hoholds and NPISH's (S.1M), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts. Sum of domestc sectors Socal benefts other than socal transfers n knd (D.62) Informaton comng from quarterly general government accounts and from the Bop are d. No quarterly nformaton relatng to Belgan prvate socal schemes s avalable and smoothng procedu are d wth due regards to ESA2010 accountng constrants. On the s sde, quarterly estmates are bult up for the three followng tems: D.62a : all socal benefts and socal assstance benefts supported by the general government (D.62a_S.1 = D.621_S.13 + D.622_S.13 + D.623_S.13 ) D.62b : prvate funded socal benefts, excl. S.13 (D.62b_S.1 = D.622_S.12) ) D.62c : drect socal nsurance benefts from employers, excl. S.13 (D.62c_S.1 = D.623_S.1 - D.623_S.13) On the ources sde, a quarterly estmate for total socal benefts (D.62) s establshed. (A) Uses sde

40 40. Non-fnancal corporatons (S.11) D.62a: not relevant. D.62b and D.62c: the corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Fnancal corporatons (S.12) Total of D.62 s smoothed out over the four quarters of the year, by applyng a lnear trend. The decomposton between the three mentoned catego s the followng: D.62a: not relevant. D.62b: dual category: D.62b_ S.12 D.62_ S.12 D.62c_ S.12 D.62c: the followng accountng dentty s d, for each quarter : D.62c_ S12 D.612_ S12 (cf above) D.62a: quarterly estmates are drectly taken from the quarterly general government accounts. D.62b and D.62c: not relevant. Hoholds and NPISH's (S.1M) D.62a and D.62b: not relevant. For D.62c, the followng accountng dentty s d, for each quarter : D.62c_ S1M D.612_ S1M (cf above) Sum of domestc sectors. The data are calculated on the bass of Bop data. When annual data are avalable, the quarterly profle s defned by usng, as ndcator, nformaton comng from the Bop. When annual data are not yet avalable, the data comng from Bop s drectly d. Due to harmonzaton procedure between natonal accounts and Bop, the dscrepances between both statstcs for ths transacton are neglgble for the more recent years. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), general government (S.13) Hoholds and NPISH's (S.1M) Socal benefts on the s sde of the whole domestc economy [cf. supra (A)] adjusted for net socal benefts pad by the t of the world are entrely recorded as ources for hoholds and NPISH's. For each quarter : D.62_ S1M D.62_ S.1 D.62_ S.2 D.62_ S.2 Sum of domestc sectors (here lmted to S.1M).

41 41. The data are calculated on the bass of Bop data. When annual data are avalable, the quarterly profle s defned by usng, as ndcator, nformaton comng from the Bop. When annual data are not yet avalable, the data comng from Bop s drectly d. Due to harmonzaton procedure between natonal accounts and Bop, the dscrepances between both statstcs for ths transacton are neglgble for the more recent years Socal transfers n knd (D.63) Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) Data are drectly taken from the quarterly general government accounts set up at t+85 days. Quarterly socal transfers n knd are equal to quarterly ndvdual consumpton expendture by the general government (P.31_S.13). Hoholds and NPISH's (S.1M) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Sum of domestc sectors. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), general government (S.13), t of the world (S.2) Hoholds and NPISH's (S.1M) Quarterly socal transfers n knd pad by the government and hoholds and NPISH's entrely beneft the latter. For each quarter : D.63_ S1M D.63_ S.13 D.63_ S.1M Sum of domestc sectors (here lmted to S.1M).

42 Other current transfers (D.7) Net non-lfe nsurance premums (D.71) No specfc quarterly nformaton s avalable relatng to net non-lfe nsurance premums. Smoothng methods are d. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts. Hoholds and NPISH's (S.1M) Ths s the dual tem for transacton D.71. For each quarter : D.71_ S.1M D.71_ S.1 D.71_ S.11 D.71_ S.12 D.71_ S13 Quarterly non-lfe nsurance premums pad by the total economy are defned accordngly to quarterly non-lfe nsurance receved by the domestc sectors adjusted for net corpondng flows wth the t of the world. For each quarter : D.71_ S.1 D.71_ S.1 D.71_ S.2 D.71_ S.2 (B) Resources sde Non-fnancal corporatons (S.11), general government (S.13), hoholds and NPISH's (S.1M) Fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Sum of the domestc sectors (here lmted to S.12) Non-lfe nsurance clams (D.72) No specfc quarterly nformaton s avalable relatng to non-lfe nsurance clams. Smoothng methods are d.

43 43. (A) Uses sde Non-fnancal corporatons (S.11), general government (S.13), hoholds and NPISH's (S.1M) Fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Sum of the domestc sectors (here lmted to S.12). (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts. Hoholds and NPISH's (S.1M) Ths s the dual sector for transacton D.72. For each quarter : D.72_ S.1M D.72_ S.1 D.72_ S.11 D.72_ S.12 D.72_ S13 Quarterly non-lfe nsurance clams receved by the total economy are defned accordngly to quarterly non-lfe nsurance clams pad by the domestc sectors adjusted for net corpondng flows wth the t of the world. For each quarter : D.72_ S.1 D.72_ S.1 D.72_ S.2 D.72_ S Other current transfers (D.7N) These transfers are estmated on a quarterly bass by dstngushng current nternatonal cooperaton (D.74), mscellaneous current transfers (D.75) and VAT and GNI based EU own ources (D.76). Current nternatonal cooperaton (D.74) Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above).

44 44. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts. Sum of the domestc sectors (here lmted to S.13). Quarterly flows are the drect counterpart of current nternatonal cooperaton recepts for the general government. Ths nformaton comes drectly from quarterly general government accounts. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts. Sum of the domestc sectors (here lmted to S.13). Quarterly flows are the drect counterpart of current nternatonal cooperaton expendture by the general government. Ths nformaton comes drectly from quarterly general government accounts. Mscellaneous current transfers (D.75) Except quarterly pre-exstng data comng from the general government accounts and from the Bop (flows between the domestc economy and the t of the world), no quarterly specfc nformaton s known. Estmates must be made. To ensure a consstency of these redstrbutve flows, a complete whom-to-whom quarterly matrx s buld up. For transactons other than related to S.13 or S.2, t reles manly on smoothng procedu. A revew of the detaled methods to establsh pre-exstng data (quarterly general government transactons) that are d to buld up the QSA matrx can be found n the above-mentoned document (cf. 2.1 above). The general procedu underlyng the establshment of whom-to-whom quarterly matrces for D.75 are the followng: to avod the appearance of negatve flows, smoothng out methods of annual data are appled on nternal components of the matrces. Total on the lnes and columns are the sum of the components;

45 45. quarterly data for general government repent constrants to draw up the matrx; flows between S.1 and S.2 have to be consstent wth nformaton contaned n the Bop. A summary of the methods d to establsh the matrx, whch combnes the s sde and the ources sde, s provded below. receved by: S.11 S.12 S.13 S.1M Total S.1 S.2 S.1+S.2 from : S S.12 0 annual data smoothed wth lnear trend vertcal dual tem S13 data (*) annual data smoothed wth lnear trend annual data smoothed wth lnear trend S.13 S.13 data S.13 data 0 S.13 data S.1M 0 0 Total S.1 S.2 S.1+S.2 (vertcal) sum of domestc sectors Belgan and EUI Bop data vertcal sum (vertcal) sum of domestc sectors annual data smoothed wth lnear trend S.13 data 0 S13 data (*) vertcal sum vertcal sum annual data smoothed wth lnear trend (**) (vertcal) sum of domestc sectors Belgan Bop data+ annual data smoothed wth lnear trend vertcal sum (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors horzontal or vertcal sum Belgan Bop data S13 data (*) S.13 data Belgan Bop data + annual data smoothed wth lnear trend (***) (vertcal) sum of domestc sectors horzontal sum horzontal sum horzontal sum horzontal sum horzontal or vertcal sum (*) Guarantee fees n the context of fnancal turmol. (**) Transfers between hoholds (S.14) and NPISHs (S.15) are not consoldated. Internal transfers wthn the sector S.1M are regstered, for an equvalent amount, on the s sde and on the ources sde. (***) Transfers from ISBLSM to S2 (not ncluded n Bop data) are smoothed wth a lnear trend. Quarterly mscellaneous current transfers related to the t of the world are partally calculated on the bass of the quarterly Belgan and EUI Bop data. Other components of S.2 current transfers, n partcular cross-border transfers related to NPISH s, are not ncluded n Bop data. They are estmated by smoothng annual data wth a lnear trend. For quarterly mscellaneous current transfers related to S.13, the nformaton comes drectly from quarterly general government accounts. Other quarterly mscellaneous current transfers are estmated by smoothng annual data wth a lnear trend. VAT and GNI based EU own ources (D.76) Quarterly pre-exstng data taken from the general government accounts are d.

46 46. (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M) Data are drectly taken from the quarterly general government accounts. Sum of the domestc sectors (here lmted to S.13). Not relevant (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), general government (S.13), hoholds and NPISH's (S.1M) Quarterly flows are the drect counterpart of VAT and GNI based EU own ources pad by the general government. Ths nformaton comes drectly from quarterly general government accounts Adjustment for the change n net equty of hoholds n penson funds erves (D.8) As for socal contrbutons and benefts related to prvate penson schemes, no nformaton s for the tme beng avalable on a quarterly bass to estmate changes n the net equty of hoholds n penson funds erve and a smple smoothng method of annual data s d. (A) Uses sde Non-fnancal corporatons (S.11) The very lmted corpondng annual fgure s dvded nto four equvalent quarters. Fnancal corporatons (S.12) Quarterly estmates are based on the followng accountng dentty: The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts. Hoholds and NPISH's (S.1M), t of the world (S.2)

47 47. Nl n Belgan accounts. Sum of domestc sectors. (B) Resources sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), general government (S.13) Hoholds and NPISH's (S.1M) Hoholds are the only counterpart for ths transacton regstered on the s sde for the total economy. D.8_ S.1M D.8_ S.1 Sum of domestc sectors (here lmted to S.1M). Nl n Belgan accounts Captal transfers (D.9) Captal taxes (D.91) Quarterly pre-exstng data taken from the general government accounts are d. A revew of the detaled methods to establsh these data can be found n the above-mentoned document (cf. 2.1 above). (A) Uses sde Non-fnancal corporatons (S.11), fnancal corporatons (S.12), t of the world (S.2) Nl n Belgan accounts. Non relevant. Hoholds and NPISH's (S.1M) Hoholds are the only counterpart for ths transacton regstered on the ources sde for the general government. D.9_ S.1M D.9_ S.13 Sum of domestc sectors (here lmted to S.1M). (B) Resources sde

48 48. Non-fnancal corporatons (S.11), fnancal corporatons (S.12), hoholds and NPISH's (S.1M), t of the world (S.2) Data are drectly taken from the quarterly general government accounts. Sum of domestc sectors (here lmted to S.13) Investments grants and other captal transfers (D.9N) Except quarterly pre-exstng data comng from the general government accounts and Bop, no quarterly specfc nformaton s known. Estmates must be made. To ensure a consstency of these redstrbutve flows, a complete whom-to-whom quarterly matrx s buld up, as for quarterly estmates of D.75. It also reles manly on smoothng procedu and on the hypothess that no transactons exst between some nsttutonal sectors. The general procedu underlyng the establshment of whom-to-whom quarterly matrx for D.9N are the followng: to avod the appearance of negatve flows, smoothng out methods of annual data are appled on nternal components of the matrx. Total on the lnes and columns are the sum of the components; quarterly data for general government repent constrants to draw up the matrx; flows between S.1 and S.2 have to be consstent wth nformaton comng from Bop. A summary of the methods d to establsh the matrx, whch combnes the s sde and the ources sde, s provded below. receved by: S.11 S.12 S.13 S.1M Total S.1 S.2 S.1+S.2 from : S.11 annual data smoothed wth lnear trend vertcal dual tem S (allowng for exceptons) (*) annual data smoothed wth lnear trend S.13 S.13 data S.13 data S.13 data S.1M 0 0 Total S.1 S.2 (vertcal) sum of domestc sectors annual data smoothed wth lnear trend (vertcal) sum of domestc sectors 0 annual data smoothed wth lnear trend Vertcal dual tem annual data smoothed wth lnear trend (vertcal) sum of domestc sectors horzontal dual tem (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors (horzontal) sum of domestc sectors horzontal sum annual data smoothed wth lnear trend 0 (allowng for exceptons) (*) S.13 data annual data smoothed wth lnear trend vertcal sum horzontal sum horzontal sum horzontal sum horzontal sum horzontal or vertcal sum

49 49. S.1+S.2 vertcal sum vertcal sum S.13 data vertcal sum horzontal or vertcal sum (*) Exceptonal transactons can be regstered on the bass of ad-hoc collected nformaton. Quarterly general government accounts gve nformaton about nvestments grants and other captal transfers pad to corporatons, whether these are non-fnancal or fnancal enterprses. For the purposes of QSA, ths nformaton s allocated pro rata amongst both sub sectors accordng to the amounts pad on an annual bass. When annual accounts are not yet avalable, the most recent annual dstrbuton key s kept unchanged. Captal transfers other than captal taxes can be the ult of exceptonal transactons. Ths s notably the case for exceptonal recepts collected by the government. By constructon, the counterpart sector s non-fnancal corporatons (S.11) whch s defned as a dual tem. Specfc attenton has to be devoted to such exceptonal transactons f t appears that the counterpart sector s not S.11. Manual correctons can be brought to the usual drawng of the matrx Consumpton of fxed captal (P51C) Gven the lack of nformaton, quarterly data are estmated on the bass of a lnear trend. Snce these se are calculated wth a model usng the lfespan of past nvestments, they have a regular profle. Non-fnancal corporatons (S.11) Ths s the dual tem for transacton P51C. For each quarter : P51C _ S.11 P51C _ S.1 P51C _ S.12 P51C _ S.13 P51C _ S1M Fnancal corporatons (S.12), hoholds and NPISH's (S.1M) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts. The data come drectly from the quarterly man aggregates Acqustons less dsposals of non-fnancal non-produced assets (NP) From 2007Q1 on, data for the t of the world are drectly taken from the Bop establshed accordng to BPM6. For tme span , the corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend.

50 50. Non-fnancal corporatons (S.11) Ths s the dual tem for transacton NP. If requred, due attenton s gven to be sure that the counterpart for exceptonal transactons regstered at the level of general government s correctly recorded. For each quarter : NP _ S.11 NP _ S.1 NP _ S.12 NP _ S.13 NP _ S1M Fnancal corporatons (S.12) The corpondng annual fgure (very small amount) s smoothed out over the four quarters of the year, by applyng a lnear trend. Data are drectly taken from the quarterly general government accounts set up at t+85 days. These data take nto account specfc transactons such as sales of land and sales of moble telecom lcences (after 2005 for the latter). Hoholds and NPISH's (S.1M) The corpondng annual fgure s smoothed out over the four quarters of the year, by applyng a lnear trend. Sum of domestc sectors. The data come drectly from the Bop. From 2006 onwards, ths transacton s manly nfluenced by the development of purchases or sales of CO2 emsson rghts. The counterpart sector for these transactons s non-fnancal corporatons. The development observed n NP_S.2 s manly mputed to the sector S.11 as ths latter s consdered as the dual tem.

51 51. Annex 1 : Methodology by transacton - Overall pcture Transacton Already avalable To be calculated Comment P.1 P Lack of quarterly Supply-Use tables that would enable P2 to be calculated usng techncal coeffcents. B.1g B.1g_S.13 B.1g_S.11 B.1g_S.12 B.1g_S.1M Avalablty of quarterly estmates of value added for more than 30 branches. To splt up value added amongst the nsttutonal sectors, the annual breakdown of value added per sector, for each branch, s defned as allocaton key for quarterly data. P.31 et P.32 P.31_S.13 - P.32_S.13 P.31_S.1M P.4 P.4_S.13 P.4_S.1M - P.51 P.51_S.13 P.51_S.1M (housng) P.51_S.11 P.51_S.12 P.51_S.1M (excl. housng) P.5N P.5N_S.13 P.5N_S.1 (adjusted) P.5N_S.11 P.5N_S.12 P.5N_S.1M P.6 P.6 - P.7 P.7 - D.1 D.1_S.1 & D.1_S.13 D.1_S.2 D.1_S.2 D.21 D.21_S.1 & D.21_S.13 D.21_S.2 D.29 D.29_S.1 & D.29_S.13 D.29_S.2 D.31 D.31_S.1 & D.31_S.13 D.31_S.2 D.39 D.39_S.1 & D.39_S.13 D.39_S.2 D.1_S.11 D.1_S.12 D.1_S.1M D.1_S.1M - D.29_S.11 D.29_S.12 D.29_S.1M - D.39_S.11 D.39_S.12 D.39_S.1M The breakdown of quarterly 'prvate' gross fxed captal formaton (excl. housng) amongst S.11, S.12 and S.1M s estmated by consderng the share of each sector n the annual corpondng aggregate. If requred, quarterly data for the whole economy are adjusted for the latest quarters to ft quarterly released GDP. The breakdown per nsttutonal sector s estmated by usng smoothng methods and defnng a dual sector (S.11). A specfc methodology has been developed usng NSSO data. Snce nformaton comng from NSSO s only avalable wth a 4-month delay, the last quarter s subject to an ad-hoc estmate. Gven the lack of nformaton, the small sums nvolved and/or the relatvely regular profle of the se, quarterly data are estmated by usng smoothng methods and defnng a dual sector (S.11). Gven the lack of nformaton, the small sums nvolved and/or the relatvely regular profle of the se, quarterly data are estmated by usng smoothng methods and defnng a dual sector (S.11). QSA methodology_sec2010.doc

52 52. D.41 D.41_S.13 & complete quarterly matrx from 2003 onwards A specfc methodology has been developed on the bass of quarterly fnancal accounts and nformaton on ntet rates from dfferent sources (stock outstandng x yeld per nstrument). For quarters relatng to years before 2003, a smplfed methodology, based on smoothng procedu, s d. Intet flows are estmated before and after FISIM allocaton. D.42 D.42_S.13 & Others A specfc methodology has been developed. D.43 - A specfc methodology has been developed. D.443 D.42_S.13 Others A specfc methodology has been developed. D.45 D.45_S.13 & Others Gven the lack of nformaton, the small sums nvolved and/or the relatvely regular profle of the se, quarterly data are estmated by usng smoothng methods and defnng a dual sector (S.1M). D.441/2, D.611+D.61 3 excl. transactons for S.13 Others No quarterly specfc nformaton or ndcator s for the tme beng avalable for nsurance and prvate penson funds transactons. Quarterly data are estmated on the bass of lnear trends, wth due regard for ESA2010 accountng constrants. NSSO, D.62b, D.71, D.72, D.8 D.5 D.5_S.1 D.5_S.13 & D.5_S.1M D.5_S.11+S.12 D.5_S.2 & D.611+D.61 3 excl. prvate D.611+D.613_S.13 D611+D.613_S2 & D.5_S.11 D.5_S.12 D.611+D.613_S11 D.611+D.613_S1M The breakdown of quarterly current taxes pad by companes between non-fnancal and fnancal corporatons s estmated by consderng the share of each sector n the annual corpondng aggregate. Gven the lack of nformaton and the small amounts nvolved for S.11, quarterly data are estmated by usng smoothng methods for S.11 and defnng a dual sector (S.1M). schemes D.612 D.612_S.13 Others Hence no quarterly specfc nformaton or ndcator s avalable, quarterly data are estmated wth an ad hoc method, combnng lnear trends and ESA2010 accountng constrants. D.62 D.62a_S.13 Others Hence no quarterly specfc nformaton or ndcator s avalable, quarterly data are estmated wth an ad hoc method, combnng lnear trends and ESA2010 accountng constrants. D.63 P.31_S.13 D.74 D.74_S.13 & D.74_S.2 & By defnton, the t of the world s consdered as the only counterpart sector of general government transactons relatng to nternatonal cooperaton. D.75 D.75_S.13 & D.75_S.11 & To guarantee a consstent treatment of flows, a complete matrx s establshed manly wth the D.75_S.2 & D.75_S.12 & help S.13 data, Bop data and lnear trends. QSA methodology_sec2010.doc

53 D.91 D.91_S.13 D.91_S.1M D.9N D.9N_S.13 & D.9N_S.2 & P51C NP P51C_S.1 P51C_S.13 NP_S.13 NP_S.2 D.75_S.1M & - D.9N_S.11 & D.9N_S.12 & D.9N_S.1M & P51C_S.11 P51C_S.12 P51C_S.1M NP_S.11 NP_S.12 NP_S.1M To guarantee a consstent treatment of flows, a complete matrx s establshed manly wth the help of lnear trends, wth the excepton of specfc transactons generally mplyng the general government. In vew of the lack of nformaton, quarterly data are estmated on the bass of a lnear trend. Snce these se are calculated wth a model usng the lfespan of past nvestments, they have a regular profle. S.11 s defned as dual sector. In vew of the lack of nformaton and the small sums nvolved, quarterly data are estmated on the bass of a lnear trend, except for specfc operatons for whch nformaton s known (payment date, sectors nvolved). S.11 s defned as dual sector. 53. QSA methodology_sec2010.doc

54 54. Annex 2 : Methodology by transacton - Overall p QSA methodology_sec2010.doc