The Effect of Financial Shock on Russian Short-term Equilibrium Interest Rates

Transcription

1 The Effec of Financial Shock on Russian Shor-erm Equilibrium Ineres Raes Analyical noe Research and Forecasing Deparmen Bank of Russia November 2017

2 2 исследоа ниях The Cenral Bank of he Russian Federaion, 2017 Address 12 Neglinnaya Sree, Moscow Phone +7 (495) , +7 (495) (fax) Websie All righs reserved. The views expressed in his noe are solely hose of he auhors and do no necessarily reflec he official posiion of he Bank of Russia. The Bank of Russia assumes no responsibiliy for he conens of he noe. Any reproducion of hese maerials is permied only wih he express consen of he auhors.

3 3 исследоа ниях We esimae he impac of he Russia-specific shocks of 2014 on he shor-erm real equilibrium ineres rae. We have used wo approaches. The firs approach is based on heoreical model calculaions. The second ress on empirical esimaes based on he resuls of IMF cross-counry research on he sensiiviy of he equilibrium ineres rae o shifs in invesmen demand and supply (savings) curves. Our esimaes sugges ha he 2014 financial shock ha resriced exernal borrowing for Russian issuers riggered pp growh of he shor-erm equilibrium ineres rae. However, he economy adjused o he financial shock in hanks o he macroeconomic policy pursued: he CDS risk premium declined, invesmen resumed growh and he ne invesmen posiion gradually decreased. As a resul, he effec of he financial shock on he shor-erm equilibrium ineres rae has almos enirely vanished.

4 Inroducion The Russian economy has been under inernaional resricions for hree years; mos of hese were imposed in 2014 Q2-Q4. Resricions no aimed a individuals can be spli ino wo ypes: 1. Technological: a ban on Russian impors of dual-use goods, miliary echnologies and sophisicaed oil-field equipmen. 2. Financial: exernal borrowing limiaions (wih more han 30-day mauriy) on major, mosly sae-conrolled, Russian companies. The effec of he imposed resricions (and Russian realiaion measures) on he Russian economy is esimaed in a number of ways. All esimaes of he impac of borrowing caps are focused o some exen on GDP performance. In his paper, we have supplemened hese esimaes wih an analysis of heir impac on Russian equilibrium ineres raes. An insigh ino equilibrium ineres rae movemens is maerial for moneary policy. Firs, equilibrium ineres rae movemens affec esimaes of he effecive policy sance, which is ofen viewed as a deviaion of he acual rae from he equilibrium. Second, equilibrium ineres rae movemens may poin o a shif in he medium-erm key rae benchmark. When i comes o he equilibrium ineres rae, he real equilibrium ineres rae is usually implied. In his paper, he equilibrium rae means he ineres rae ha would sele in he economy if prices were flexible, ha is he rae where he oupu is a is poenial a any given ime. Significanly, he paper esimaes he real equilibrium ineres rae s deviaion in an economy under inernaional resricions from he equilibrium in an unresriced economy. To esimae he equilibrium ineres rae, applied research ofen links i o poenial economic growh raes. Such an approach o his noion (concep) is described in he preprin The equilibrium ineres rae: a measuremen for Russia (Bank of Russia Working Paper Series, 2016, No. 13). Theoreically, he link is linear, bu i is uneven and depends on addiional raios. I can be wrien as follows: r g ln, where r is he real rae, g is GDP growh, σ is he elasiciy of ineremporal subsiuion, and β is he discoun facor. The link beween he ineres rae and poenial growh raes is as follows. When income is likely o grow a a high pace (fuure income is higher han presen income) loans enjoy greaer demand due o consumpion smoohing. In expecaion of increased fuure income, economic agens may enhance consumpion oday hrough he use of credi. Accordingly, he higher GDP (income) growh, he higher he demand for loans o fund consumpion and, consequenly, he higher ineres raes will be, in compensaion for lenders rejecion of heir own curren consumpion in favour of borrowers increased consumpion. If poenial growh slows as i becomes clear ha he income growh rae is deceleraing (ha is, fuure income is posed o be lower han iniially expeced), economic agens adjus heir behaviour; hey sar saving more o smooh consumpion. The increase in he saving supply calls

5 5 исследоа ниях for a respecive (higher) demand for credi. In he new environmen his is possible only if ineres raes are low. The link beween he ineres rae and GDP growh depends on wo facors: 1. The so-called elasiciy of ineremporal subsiuion 1/σ. This raio reflecs he sensiiviy of consumpion (GDP, income) o ineres rae movemens. The increase in he income growh rae may be accompanied by a considerable or moderae ineres rae hike depending on he financial sysem s capabiliy o ransfer resources from lenders o borrowers. The higher he elasiciy of subsiuion, he less responsive he ineres rae is o changing income. 2. The discoun facor β. Theoreically, ln β < 0. This facor reflecs how economic agens evaluae he fuure compared o he presen. If he fuure is devalued (economic agens live for oday), he discoun facor falls considerably below 1 (he logarihm is posiive) for any income and growh pace. Thus, such impaien agens ranslae any growh in poenial GDP ino a higher ineres rae in he economy. The second iem reflecs his increase. Anoher link beween he ineres rae and GDP growh sees GDP as a produc raher han he oal expenses/income in he economy. This correlaion calls for equalisaion of he reurn on financial asses (ineres rae) and real asses (capial). r = MP + ( Pcapial Pgdp) amorisaion rae, where MP is he marginal produc of capial (faser echnological advances and an increasing labour force brings he ineres rae up whereas a slowdown decreases he ineres rae). However, his correlaion also requires adjusmens. The relaionship is no one-o-one. Paricularly, adjusmens should be made for he relaive price of invesmen goods ( Pcapial Pgdp). Is reducion favours real raher han financial invesmen and raises he ineres rae. The increase in he amorisaion rae reduces gains from physical capial invesmen (because of resale complicaions). The raio of GDP growh o ineres rae growh is also valuable in esimaing he sabiliy of he economic equilibrium, ha is, an esimae of he non-explosive naure of deb increase in he economy. For his purpose, he real ineres rae reading should exceed GDP growh (r > g). This is he so-called modified golden rule. Deb growh is where d is real deb, r is he real ineres rae, g is GDP growh and p is inflaion., The rule says ha when GDP (income) growh surpasses real ineres raes (r < g), economic agens may borrow easily. They accumulae deb a low raes and, imporanly, may borrow o pay ineres on loans ( a snowball effec ) bu his fails o increase he deb-o-income raio as he deb is dilued faser han capialised ineres accumulaes. The equilibrium is possible if r < g, bu i is ineffecive as economic agens may always ake advanage of new opporuniies o accumulae deb. As a resul, he equilibrium shifs o r > g sooner or laer. When r > g, such conduc proves impossible, deb will snowball and, hus, for he equilibrium o be sable, he condiion r > g calls for a deb finieness (no higher han a cerain value in erms of income). If r > g, financial bubbles may arise.

6 6 исследоа ниях In effec, such esimaes prove o be very uncerain, as shown by sudies of correlaion esimaes beween he equilibrium rae and he poenial economic growh rae (have large confidence inervals). This poin should be aken ino accoun in pracical applicaion of he esimaes (e.g. moneary policy advice). I should be also borne in mind ha exernal borrowing complexiies may have a mixed effec on real equilibrium ineres raes. They impac boh invesmen supply and demand. On he one hand, resriced access o new exernal borrowing reduces accessible financial resources a home, hus decreasing he supply of financial resources available for invesmen. Furhermore, rising risks relaed o resricions, along wih economic uncerainy, may curail he inflow of nonresiden funds o he indusries and secors unaffeced by inernaional sancions. Boh hese facors shif he savings curve S lefwards and upwards (Figure 1). If he invesmen demand curve remains unchanged, his should increase he equilibrium rae and reduce he equilibrium invesmen volume. On he oher hand, a decrease in companies demand for invesmen shifs he I curve lefwards and downwards. This leads o a drop in he equilibrium value of he ineres rae and invesmen volume. The decline in demand may be caused by heighened uncerainy, he emergence of free producion capaciies following GDP conracion, reduced accessibiliy of echnologies (and, consequenly, lower effeciveness of invesmen projecs), changes in he relaive prices of invesmen goods, 1 ec. Figure 1. Invesmen demand and saving supply movemens r S 1 S E 1 E 0 I 1 I The resuling change in he real equilibrium ineres rae is ambiguous, as i depends on he shif of he invesmen demand (I) and saving supply (S) curve. A he same ime, invesmens in he economy are clearly shrinking. We applied wo approaches o esimae he effec of exernal financing resricions on equilibrium ineres raes: calibraed heoreical model calculaions and IMF research-based empirical esimaes. S, I 1 The ruble s depreciaion led o a more considerable growh in he cos of impored machinery, equipmen and maerials in ruble erms compared wih he increase in prices of domesic inermediae and invesmen goods and exfacory prices.

7 7 исследоа ниях Financial marke indicaors allow us o esimae shor-erm equilibrium rae movemens hrough changes in risk premiums. Thus, in he period beween lae 2014 and early 2015, he Russian risk premium (5-year CDS premium) rose by 4.7 pp while risk premiums of oher emerging markes hardly changed. A he same ime, he risk premium is highly volaile, complicaing is use in esimaes of shor-erm equilibrium rae movemens. Empirical calculaion resuls We used he findings of an IMF sudy 2 o esimae equilibrium ineres rae movemens. I assesses he effec of invesmen and saving curve shifs on he equilibrium ineres rae and invesmen/savings. In is research he IMF esimaed he following sysem of simulaneous equaions on global economic daa from 1980 o 2013: s i s a b i 0 0 a r a n 1 b r b p where S denoes saving as a per cen of GDP, i is invesmen as a per cen of GDP, r represens he real ineres rae, n is advanced economy social expendiure as a per cen of GDP 3 and p sands for he relaive price of invesmen calculaed as he change from invesmen deflaor o GDP deflaor. The research did no consider resricions on exernal borrowing in any economy; hus, we esimaed heir effec hrough he shif in saving and invesmen curves and heir impac on he equilibrium ineres rae. We assumed ha he saving curve shif equalled conracion in foreign deb (forced deleveraging) while he shif in he invesmen demand curve equalled he change in he relaive price 4 of invesmen. 5 Significanly, he imposiion of inernaional consrains or bans on raising capial in foreign markes for Russian companies coincided wih a major drop in oil prices. Shrinking oil and gas expor revenues could also have a negaive effec on domesic saving. However, as he exchange rae rapidly adjused o he changes in he balance of paymens, impors conraced considerably. Is scale was comparable wih he drop in expor coss and, hus, he effec of he oil price downurn on funds available for invesmen proved o be neural. Moreover, he ruble s depreciaion increased coss of invesmen equipmen and maerials. Our esimaes sugges ha, adjused for commodiy prices, he ruble would have depreciaed by only 25-35% compared wih he acual 56.5% weakening (afer he ruble appreciaion in 2015 Q2). However, we calculaed he effec of he acual ruble weakening on invesmen demand as an upper bound esimae. 2 Recovery Srenghens, Remains Uneven // WEO. Chaper April. 3 Given ha increased social expendiure allows households o save less, all else being equal. 4 We calculaed he change in he relaive price of invesmen hrough he relaive change in fixed capial invesmen deflaor and GDP deflaor. 5 Ineresingly, in he IMF sudy he effec of he change in he relaive price of invesmen has a posiive sign, which is couner-inuiive a firs approximaion. The increase in he relaive cos of invesmen goods reduces reurn on invesmens, lowering demand for invesmens (he quaniy of purchased equipmen) a a given ineres rae (a lefward shif in he invesmen demand curve). On he oher hand, he increase in he invesmen cos implies ha he purchase of one equipmen iem will require a larger nominal loan a a given ineres rae (a righward shif in he invesmen curve). The shif in he demand curve depends on which facor is of higher imporance for he invesmen demand. The IMF sudy revealed ha he elasiciy of invesmen demand o he relaive price of invesmen goods proved posiive for he global economy in he period since Tha is, price increases shif he invesmen demand curve o he righ, raising he equilibrium ineres rae, all else being equal.

8 US dollars, bln Analyical noe 8 исследоа ниях We used he IMF s raios of he equilibrium rae s sensiiviy o shifs in saving and invesmen curves o obain he following resuls: 1. The real equilibrium ineres rae rose by 0.9 pp by he end of he firs quarer of Exernal asses (ne posiive inernaional invesmen posiion) helped miigae he effec of financial resricions on he equilibrium ineres rae. The shrinkage of exernal asses ousripped exernal liabiliies, 6 serving as a buffer for he Russian economy. Wihou hem, he real equilibrium ineres rae would have grown by 9 pp agains he 2014 Q2 reading. 3. The Bank of Russia s moves played a considerable role in miigaing he effec of exernal borrowing resricions on equilibrium ineres raes (Figure 3). I was he Bank of Russia s FX liquidiy provision (paricularly, FX repos) ha largely reduced he effec of he financial channel of inernaional resricions on he shor-erm real equilibrium ineres rae, especially in he fourh quarer of The surge in he equilibrium ineres rae proved emporary. Despie ongoing resricions on exernal borrowing, we may conclude ha he economy has almos compleely adaped o he financial shock. Thus, he ne inernaional posiion has declined progressively since early 2016 due o greaer growh in exernal liabiliies han in asses. This poins o he growing accessibiliy of exernal financing. Furhermore, he policy of he Bank of Russia and he Minisry of Finance resored invesor confidence, as confirmed by he decline in he Russian CDS risk premium hroughou 2016 and is reurn o he 2013 H2 readings in early Figure 3. ey ne inernaional invesmen posiion componens, $ bln Liabiliies Oher asses Reserve asses Ne Inernaional Invesmen Posiion IIIQ 2014 IVQ 2014 IQ 2015 Source: Bank of Russia, R&F Deparmen calculaions. Figure 4. Russian CDS (Russia CDS USD SR 5Y Corp) Source: Bloomberg Finance L.P. Model calculaion resuls In order o esimae he likely effecs of inernaional financial resricions on he Russian economy, we consider a simplified hree-period model. 7 I implies ha, in he firs period, economic agens learn ha hey should compleely redeem heir deb in he second period, and 6 Adjused for revaluaion. 7 Our calibraion suggess ha one period corresponds o one year.

9 9 исследоа ниях reach a specific capial value by he hird period. 8 We have compared he resuls obained in his model wih hose of he model where he deb is refinanced in full. Thus, we find he effec of forced deleveraging (exernal deb reducion), which is idenical o he effec of exernal finance resricion in he model. The indicaors were calibraed based on Russian specificiies. For example, he deb-o-oupu raio roughly equals he raio beween ne privae exernal deb o he Russian GDP before he imposiion of financial resricions. The iniial calculaion was based on a relaively siff assumpion ha he producive capial should reurn o he iniial level in he hird period of ime. Figure 5 feaures he respecive resuls. The main effec on he ineres rae falls on he hird period (ha is, he hird year). The need o fully repay deb reduces he economy s capial sock and, in order o replenish i o he iniial level in he hird period, condiions mus be esablished o promply shif consumpion owards invesmen. To do so, a srong rae increase is needed o make saving more aracive. Figure 6 feaures he resuls, where he ulimae capial sock ranges beween 90% and 100% of he iniial value. The model provides for a reurn o he saionary level in he fuure. The weakening of he assumpion of he level o which capial should reurn in he hird period is enirely logical. Indeed, as long as, for example, a ban or siff resricions on cerain echnology impors are effecive or a long-erm equilibrium oil price persisenly lowers, he poenial oupu and, consequenly, he required capial fall compared wih he iniial level. The effec o be reached in he hird period is very sensiive o assumpions abou he level of capial o be reached by he end of his period. Moreover, in he second period of ime, he effec ranges beween 0.55 and 1.36 pp when capial changes is level from 90% o 100% of he iniial reading. We have used i o gauge he effec of financial resricions on ineres raes, as we are primarily ineresed in he impac of he forced deleveraging on he equilibrium rae. Figure 5. Rae hike afer deb redempion in he scenario where capial reurns o he iniial reading, pp Figure 6. Rae hike afer deb redempion depending on he capial level in he hird period, pp 8 For model deails please refer o he Appendix.

10 10 исследоа ниях % 92% 94% 96% 98% 100% The heoreical model calculaions are overall in line wih he empirical esimaes. The use of forced deleveraging o resore he iniial capial reading and non-use of accumulaed exernal asses o se off he closed funding channels would cause ineres raes o jump (by 17 pp, according o model esimaes). The empirical esimaes ha do no cap he capial level o be reached in he economy sugges ha raes could increase by roughly 9 pp during he course of he year. The esimaed effec of financial resricions on he shor-erm real equilibrium ineres rae declined considerably, if we assume ha economic agens have failed o compleely resore he iniial capial level and ake he ne posiive inernaional invesmen posiion ino accoun. The laer permied he use of a par of exernal asses o offse he closure of he exernal deb financing channel. The resuling effec was pp according o model esimaes and 0.9 pp according o empirical esimaes. Conclusion Boh calculaions sugges ha he imposed resricions on raising new exernal financing by Russian companies increased shor-erm real equilibrium ineres raes by roughly 1 pp. This increase is quie subsanial and suggess ha he Bank of Russia s moderaely igh moneary policy of was jusified. The consisency of he Bank of Russia s policy during ha period, among oher hings, reduced he risk premium from is peak value of almos 600 bp in lae 2014 o bp by lae 2015, despie persisenly falling oil prices. Laer, he recovery of oil prices in 2016 alongside he moderaely igh moneary policy led o furher reducion in he risk premium o he readings regisered before he shock. As a resul, he effec of he rise in he equilibrium ineres rae, riggered by he financial shock, was largely offse and has almos compleely vanished.

11 11 исследоа ниях Appendix No-financial-resricion model We consider a simple hree-period model of a small open economy where agens maximise heir uiliy. l l l, Given he resricions: ( r) ( ) ( r) ( ) ( r) ( ), where is consumpion in he period of ime, is capial in he period of ime, is deb in he period of ime. The firs order condiions will be wrien as follows: a a ( ) ( ) ( r) ( r). Le us assume ha The iniial and he ulimae saes are as follows: ( r). ( a ( ) ) I is clear ha: r

12 12 исследоа ниях for any. Financial resricion model Similar o he basic model, economic agens maximise heir uiliy: l l l, Given he resricions: ( r) ( ) ( r) ( ) ( ) ( ( ) ). The firs order condiions will be wrien as follows: a a ( ) ( ) ( r). Calculaing raes in he model The model conains no raes; hus, we calculae hem as follows: Calibraion Parameer. Value -0.3 a

13 13 исследоа ниях References Drehmann M., Juselius M. Do Deb Service Coss Affec Macroeconomic and Financial Sabiliy? // BIS Quarerly Review Sepember 1. Available a: hps://ssrn.com/absrac= Sufi A. Ou of Many, One? Household Deb, Redisribuion, and Moneary Policy During he Economic Slump // Andrew Crocke Memorial Lecure June. Dones S., Ponomarenko A., Measuring Deb Burden // Bank of Russia Working Paper Series No. 5. repsev D., Porshakov A., Seleznev S., Sinyakov A., The equilibrium ineres rae: a measuremen for Russia // Bank of Russia Working Paper Series No. 13. Research and Forecasing Deparmen Yulia Ushakova Dmiry Chernyadyev The auhors express heir graiude o Sergey Seleznev for his conribuion o he heoreical model.