Surprise: Riksbank is now your SEK friend

Transcription

1 Surprise: Riksbank is now your SEK friend FX Survey and SEK outlook: Riksbank now a krona positive. Our survey of major SEK trading companies and institutions has ranked the Riksbank as the major negative factor for the SEK since (Nov 2013). However, the latest survey shows perceptions regarding future monetary policy have changed with the Swedish central bank now seen as a clear positive for the krona. FX interventions are highly unlikely according to 92% of respondents. Despite these more hawkish views, the market still holds a fairly limited krona position (+9%). Our call since H for cautious and continued trade-weighted SEK appreciation remains and is supported by the survey. Swedish macro and Riksbank outlook: Strong growth risks a taper tantrum. We forecast Swedish GDP will grow by 4.0% in 2016, partly due to an expected acceleration in public consumption and investment driven by immigration inflows last year. In our view, international growth concerns have been exaggerated and the current positive trend for the Swedish industrial sector will continue. We forecast an end to Swedish rate cuts and QE programme: we see a risk of a similar, albeit smaller market reaction compared to the US taper tantrum in If we are correct the yield curve will steepen while the krona should appreciate. Service inflation: Key to raising CPIF to 2% y/y. Swedish service inflation accelerated to 2% y/y in January this year from less than 1% y/y at the end of Such inflation must increase significantly before CPIF is likely to revert to 2% y/y in an environment characterised by low international price pressure and decreasing support from a weak exchange rate. Budget outlook: Not as rosy as you may think. Swedish government finances have surprised on the upside. But because they are highly sensitive to the business cycle, changed assumptions on growth or interest rates could significantly alter the outlook going forward. Sweden and Brexit: How to position if UK votes to leave EU. We would expect a decision by the UK to leave the EU to increase uncertainty, lower growth and push both yields and the SEK lower. However, our survey respondents believe EUR/SEK will decline in a Brexit. Government bonds & FX: We expect the Riksbank to remain on hold at its monetary policy meeting on April 21 and to keep both the repo rate/rate path unchanged. We do not forecast any new QE measures. We favour the following Swedish fixed income trades: sell Dec 16 RIBA future; position for a flatter 5y/10y SEK IRS curve; and receive 5y5y SEK vs. EUR IRS. In FX we see value in selling USD/SEK calls at 8.40/50, buy GBP/SEK calls and hold on to short EUR/SEK (entry 9.40). TUESDAY 12 APRIL 2016 EDITOR Carl Hammer Historical SEK positioning 20 Market ranking of SEK drivers 60% Overweight in SEK 60% 15 40% 40% 10 20% 20% 5 0% 0% 0-20% -20% -5-40% Underweight in SEK -40% -10 Own company Market

2 Forecasts CENTRAL BANKS EMU US UK SWEDEN NORWAY Current target rate 0.00% % 0.50% -0.50% 0.50% Apr-16 Unch Unch Unch Unch --- May Unch --- Unch Jun-16 Unch Unch Unch --- Unch Jul-16 Unch Unch Unch Unch --- Aug Unch --- Unch Sep-16 Unch +25 bps Unch Unch -25 bps Oct-16 Unch --- Unch Unch --- Nov Unch Unch --- Unch Dec-16 Unch +25 bps 0.50% Unch Unch End % % 1.00% 0.50% 0.25% CONTENT SEK FX survey & Flow outlook. Page 3 SEK is long-term undervalued. Page 7 Swedish Macro & Riksbank outlook. Page 9 Service inflation: Key to raise CPIF. Page 14 Ahead of fiscal spring budget. Page 16 Sweden, SEK & Brexit. Page 19 Swedish Fixed Income trade ideas. Page 21 Disclaimer. Page 23 CONTACTS Carl Hammer, Head of FX Research (Editor) Jussi Hiljanen, Head of Fixed Income Research Daniel Bergvall, European Economist Olle Holmgren, Senior Swedish Economist Richard Falkenhäll, Senior FX Strategist FX FORECASTS FORECASTS ACTION LEVELS FX 11/04/2016 1m Q2 16 Q3 16 Q4 16 Q1 17 Q2 17 Q3 17 Q4 17 Buy Sell EUR/USD EUR/SEK EUR/NOK EUR/CHF USD/JPY EUR/GBP USD/SEK GBP/SEK NOK/SEK JPY/SEK CNY/SEK CHF/SEK *NOK Forecasts are subject to revision on Apr 13th when SEB releases NOK Views. 2

3 FX survey & SEK outlook: Riksbank is now on your side Our survey of major companies and institutions trading SEK shows the market currently sees major changes ahead for the Riksbank. For the first time in a long while, the central bank is regarded as a positive factor for the krona, with its members increasingly tolerant of the present inflation undershoot and likely to prioritise objectives other than just low spot inflation. However, market positioning and projections for the SEK will likely remain fairly cautious as survey participants do not expect a full scale reversal of monetary policy anytime soon. Since 1999, SEB has published SEK Views to better understand market positioning and the outlook for the currency, by polling major SEK-trading companies and institutions. Since H2 2015, sentiment has steadily improved after a period of 2-3 years during which both investors and companies said they preferred to underweight the currency. In our latest survey overall positioning remains small long SEK (+9%) driven by a further improvement from the corporate sector. While participants believe the overall market holds its longest position in three years (at +15%), the levels identified are fairly cautious and neutral for the outlook for the exchange rate. 60% Historical SEK positioning Overweight in SEK 60% CONCLUSION ON POSITIONING: Somewhat surprisingly participants remain (only) cautiously overweigh to the SEK, while sentiment has not improved since our last survey. Positioning is therefore neutral for the currency. SUBSTANTIAL CHANGE IN KRONA DRIVERS. Central banks have been at the forefront in driving exchange rates in recent years. Indeed, when we asked our survey participants to rank different krona drivers, results show they have regarded ECB monetary policy as very positive for the SEK. Conversely, the Riksbank has been regarded as a very negative and hostile factor for the SEK (our last survey showed ECB +22 vs. Riksbank - 30). Today however, respondents perceptions have changed quite considerably: while ECB policy remains SEK positive (at +15 = not surprising) so also is the outlook for Swedish monetary policy and its effect on the SEK (+10 now). This result is also consistent (see below) with the market s present views on Riksbank policy. While other drivers are deemed unimportant, as usual risk appetite is the SEK s worst enemy. This last finding is also surprising as the SEK is currently uncorrelated to risk appetite (see further details below). 20 Market ranking of SEK drivers % 40% 20% 0% -20% -40% Underweight in SEK Own company Market 20% 0% -20% -40% A MORE HAWKISH RIKSBANK ANTICIPATED. Recent signals from the Riksbank (and indeed market pricing) that potentially its policy easing is coming to an end have obviously been picked up by the market, with 67% of respondents expecting that monetary policy will not be eased any further. If instead the Swedish central bank does choose to ease policy yet again, survey participants favour an extension of QE (69%). Participants' own SEK positioning 68.0% Will the Riksbank be more tolerant to inflation under target? 86.5% 30.0% 30.0% 24.0% 40.0% 8.0% 13.5% Overweight Neutral Underweight Yes No Corporates Financials 3

4 Perhaps a key element of the change in the reaction function of the Riksbank is that it is now perceived as being more tolerant of inflation undershooting its target with 86.5% of respondents agreeing with this proposition. Further, several board members are now more sure that the Swedish economy is sufficiently strong and resource utilisation tight enough for inflation to be set to return to its 2% target. The market is also convinced that actual FX interventions are highly unlikely as almost 92% of respondents do not expect them to occur. Those few who see interventions coming say that a level just above 9.00 in EUR/SEK represents the current floor for the currency pair. Riksbank to intervene in the FX market? 92.3% 7.7% UK vote to leave EU & EUR/SEK 43.6% 33.3% 15.4% 0.0% STRONGER USD, EUR/SEK IS STILL OVERVALUED Market forecasts are in line with the consensus estimate of a slightly lower EUR/SEK and slightly higher USD/SEK. Unsurprisingly, the USD is still expected to be the strongest currency with 53% of respondents saying EUR/SEK is overvalued (34% disagree and regard the EUR as undervalued vs. the SEK at its current rate). No 7.7% Yes As regards other central banks, only 22% of survey respondents believe the ECB will not announce new policy measures. Participants are divided between expecting the Fed to hike rates in June (32%) and September (43%). CONCLUSION ON SEK DRIVERS. Survey participants have reversed their previous views on the Riksbank policy outlook and its effect on SEK developments. As shown above, this outcome is logical given how the market foresees future Swedish monetary policy reaction function. EUR/SEK USD/SEK Jun-16 Dec-16 Jun-16 Dec-16 Average Median High Low Average during survey EUR/SEK 9.27, USD/SEK 8.14 Implied EUR/USD fc: 1.11 and % Which currency will outperform the others in the upcoming 6 months? 36% 10% 10% SEK DEEMED DEFENSIVE SHOULD BREXIT OCCUR. The UK will vote on whether to stay or leave the European Union on June 23 (a No vote can/will likely be followed by renegotiations on UK membership). Sterling has already weakened substantially as the market sees considerable uncertainty regarding the outcome to the referendum (SEB expects the UK to vote Yes). We asked survey participants what they thought market reactions would be to a potential No vote. In reply they said: 1) the effect on Swedish growth would be to weaken it (42% vs. unchanged 47%); 2) Riksbank policy would be more expansive (34% vs. unchanged 58%) and; 3) EUR/SEK would weaken if the UK votes to leave the EU (net 28%). In a separate article To Brexit or not Brexit we analyse the effects of such a decision on Sweden, the SEK and Swedish interest rates. USD SEK NOK EUR CONCLUSION SEK VIEWS SURVEY. The market is cautiously positioned and has relatively low expectations on future SEK developments. Most importantly however, the Riksbank is now regarded as focused on strong economic developments and their effect on future inflation. If this proves correct we should also expect a stronger krona over time. 4

5 FLOWS ANALYSIS: FOREIGN SELLING, DOMESTICS BUYING! Sweden generates a 7%/GDP current account surplus. As stated in previous SEK Views, the surplus is changing both its character and composition as: 1) the traditional trade surplus has been in structural decline since early Of the SEK 121bn surplus posted in 2015, some SEK 80bn was derived from merchanting activities (i.e. when Swedish companies buy and sell goods abroad). These are unlikely to generate FX flow as the positive surplus is partly offset by reinvested earnings abroad; 2) The services trade surplus is steadily increasing, rising to SEK 107bn in 2015 from SEK 72bn in 2014, and; 3) Net investment income is also in relatively large surplus. In the context of very large international capital flows, the SEK bn surplus (after reinvested earnings) each year is SEK positive but of fairly limited importance for 3-6 month krona developments. Sweden is a large net creditor in equity denominated investments (FDI primarily) but also a debtor in fixed income-related investments. The net investment position regarding portfolio investments is a relatively steep -50%/GDP. Overall the Swedish net international investment position (IIP) is balanced (0%/GDP), although several indications show the country should possess a sizable surplus based on accumulated current account surpluses and net errors and omissions. net selling foreign shares and funds (inflow: SEK 78bn). SEK DRIVERS GOING FORWARD Since the financial crisis in , domestic investors and companies have raised and maintained a historically high level of FX exposure (low hedging). Meanwhile our speculative index has shown foreign institutions to be net SEK sellers (see graph below). But from Q when the Riksbank also initiated its nonconventional monetary policy (and the trade-weighted SEK bottomed out) speculative investors switched to become net SEK buyers. However, since December 2015, harsh rhetoric from the Swedish central bank and threats of actual FX interventions have caused speculative flows to turn lower once again. But clearly foreign speculative investors will be SEK buyers once the Riksbank deviates from the ECB s dovish policy. Swedish domestic companies and institutions remain underweight to the krona and will ultimately lower their FX exposure. There are therefore still good reasons why the trade-weighted SEK will continue higher over the next 12 months. SEB net SEK speculative flows (weekly sentiment indicator) Most headlines concerning portfolio flows have focused on the declining foreign ownership rate in 2015 (see graph above). Coinciding with the advent of Riksbank Swedish government bond purchases, foreign investors began scaling back their investments in such bonds. Statistics Sweden data are hard to interpret on this issue (while the government authority is investigating portfolio flows): overall net portfolio flows in long-term Swedish government bonds (denominated in SEK) increased by SEK 37bn in 2015 while nonresidential holdings of Swedish interest-bearing securities fell by nearly SEK 140bn last year, almost fully matching purchases made by the Riksbank since the QE program was introduced. Total net portfolio flows 2015 showed a positive inflow of SEK 87bn, though this was primarily the result of Swedish investors DIVIDEND SEASON: LIMITED NEGATIVE SEK EFFECT. Swedish companies distribute annual dividends between March and May. The following graph charts weekly distribution and assumes for simplicity a fixed foreign ownership rate of 40% and a reinvestment rate of 80%. Of course the degree of foreign ownership varies considerably and individual investors will select reinvestment rates of between 0% and 100% (although historically 80% is normal). Using rates of 40% and 80% respectively yields a meager payout equal to 8% of the total dividend pool for conversion into FX, or just under SEK 10bn of the total SEK 120bn payout. This is a small (and negative) SEK flow compared to total turnover. 5

6 As regards seasonality during the dividend distribution periods we observe that EUR/SEK has tended to depreciate at the start of the year: between 2011 and 2015 on average Q1 has seen a 1.4% (accumulated) decline, though in Q2 (when most dividends are distributed) EUR/SEK normally rises by an accumulated 0.8%. For USD/SEK the pattern is highly inconclusive. For KIX (trade-weighted SEK) the historical pattern also shows a slightly weaker SEK (1-1.5% during Q2). So without drawing any firm statistical conclusions the SEK tends to depreciate slightly during the equity dividend season. DOVISH FED TO CHANGE USD/SEK AND EQUITY CORRELATION? Swedish financial institutions maintain a historically high FX exposure. When the Fed announced its upcoming rate hike cycle in 2014, the USD outperformed all other currencies while at the same time asset markets continued to perform. This altered the previous longterm negative correlation between the USD and equities. For asset managers this meant a strong SEKdenominated portfolio performance. However, the Fed s recent dovish turn is currently changing this correlation. Also, USD/SEK is still overvalued, which likely makes profit-taking tempting for investors long of US assets where the currency is unhedged. Hence the risk for a too strong currency for comfort for Riksbank remains primarily related to a weaker USD (rather than a dovish ECB). CONCLUSION SEK OUTLOOK. Obviously the signals sent by Riksbank as of late (inertia to add more stimuli) has been pick-up by the market. Riksbank is regarded as a positive factor for SEK and the central bank is expected to be more tolerant to current low spot inflation. The market is cautiously positioned in SEK and as long as Riksbank doesn t hint of a policy reversal krona appreciation will be very slow and orderly. Foreigners have cut their Fixed income exposure to Sweden and/but Swedish investors have repatriated capital abroad as risk appetite has turned lower. And Swedish investors and corporates remain with a high FX exposure. Hence there are a few factors which speak for SEK not being the very procyclical currency we saw prior/during the financial crisis Strong growth, tighter resource utilization and less Riksbank focus on ECB will underpin more SEK appreciation as does current positioning and SEK valuation. 6

7 The krona is long-term undervalued Long term trends exist in the nominal exchange rates of almost all currencies. However, this is not the case with the krona. Instead the real trade weighted krona index includes a long-term downward trend. This signals that Swedish competitiveness is most likely substantially stronger today than in 1995 or put differently that the krona is undervalued in nominal terms. This is in line with the findings of our long-term fair value model, which indicates the krona is 13% below its long-term fair value against the dollar and 10% against the euro. THE KRONA BEHAVES DIFFERENTLY. Long term trends exist in the nominal exchange rates of almost all currencies, and even for nominal trade weighted currency indices. For most of the time, these long-term trends in nominal exchange rates are related to differences in relative inflation rates between countries. Over a period these tend to be offset by changes in the nominal exchange rate according to the purchasing power parity theory. So a country with a lower inflation rate relative to others will usually see its currency appreciate to maintain purchasing power parity. However, this is not the case with the krona. Instead the real trade weighted krona index includes a long-term downward trend, as a result of which the krona is today unreasonably weak given the inflation differential between Sweden and its trading partners. Indeed, the krona is at precisely the same level in nominal terms against the country s trade counterparts today as it was at the beginning of 1995, just two years after it started to float. However, since then Swedish CPI inflation has been lower than in these countries, improving Sweden s competitiveness. Or conversely, given the difference in inflation between Sweden and its trading partners the krona should have appreciated by 16% in nominal terms to maintain the same real trade weighted exchange rate that prevailed in Clearly, this signals that Swedish competitiveness is most likely substantially stronger today than in However, it is reasonable to argue that CPI inflation may not be the best way to measure a country s competitiveness. More relevant might be measuring production expenses including the unit labour cost (ULC), which captures developments in both wage costs and productivity growth. The IMF calculates real effective exchange rates based on relative ULC between countries. Using this kind of deflator the argument that the krona is substantially undervalued compared to Sweden s most important trading partners, improving Swedish competitiveness, becomes even more persuasive. The index for the real exchange rate deflated by relative ULC has decreased to around 60 from 100 in 1995, more than the measure discussed earlier. This provides an even more convincing illustration that Swedish competitiveness must have increased over time as the improvement in relative ULC has not been offset by a similar appreciation of the krona. To revert to the same level of competitiveness that existed in 1995, the nominal krona trade weighted index would need to appreciate by almost 40% from its present level. Apparently therefore, the current exchange rate policy has successfully improved Swedish competitiveness over a considerable period. Despite the country experiencing lower inflation than its trading partners and a stronger development in relative unit labour costs, the krona is basically unchanged in nominal trade weighted terms. However, this is kind of difficult to understand given that the currency has been floating since 1992 and for most of the time has not been officially managed by either the government or the Riksbank. How do we explain this situation? One factor that may perhaps partly do so is the negative Swedish terms of trade development. Since 1995, Swedish terms of trade have fallen by around 10% as prices of Swedish export goods have declined compared to prices for the country s imported goods. Of course, 7

8 that certainly does not fully explain the improvement in Swedish competitiveness. A second possible explanation lies in suggestions that Swedish exporters apparently hold profits abroad, restricting capital inflows despite a surplus in the trade and current account. Nevertheless, the conclusion still holds that the krona most likely is undervalued in nominal terms today given the development in relative CPI and ULC supporting Swedish competitiveness. KRONA UNDERVALUED ACCORDING TO SEBEER- MODEL. For a long time, we have argued that the krona is cheap against most currencies and in trade weighted terms. The same view is supported by our internal longterm fair value model (SEBEER), which suggests that the long-term fair value in EUR/SEK is around 8.50 and 7.20 in USD/SEK. If these estimates are correct it would indicate there is currently one to two standard deviations by the krona from its estimated equilibrium exchange rates against these two major currencies, providing further evidence that it is in fact undervalued. Although the Swedish currency trades near its post-1993 historical average against the euro, we would argue that the krona is undervalued and that nominal exchange rates are likely to appreciate over the longer term against most currencies including the euro and the dollar. This is particularly the case as Sweden appears to have benefited from a significant improvement in competitiveness since the krona began to float in During this period, Swedish CPI-inflation and the ULC compared to other countries appear to suggest the nominal krona exchange rate should be stronger against most currencies, while instead it is basically flat or even weaker than in 1995, a conclusion which is also supported by our internal long-term fair value model. 8

9 Swedish GDP growth to remain firm in 2016 We forecast Swedish GDP will grow by 4.0% in 2016, partly due to an expected acceleration in public consumption and investment driven by refugee and immigration inflows last year. In our view, international growth concerns have been exaggerated and the current positive trend for the Swedish industrial sector will continue. GROWTH WILL REMAIN STRONG IN Following a decline in Swedish sentiment indicators at the beginning of this year several forecasters have lowered their 2016 GDP estimates. For example, in March, the NIER cut its 2016 GDP estimate to 3.5% from 3.8% in December. The NIER forecast is now in line with that of the Swedish central bank. However, our analysis in the Riksbank section shows we expect the bank to raise its 2016 GDP estimate. One reason for this assessment is that GDP growth in Q4 exceeded the Riksbank s expectations by almost a full percentage point and its higher level at the end of last year would all else being equal add 0.5%- points to average 2016 growth. The central bank could rightfully argue that quarterly national accounts are volatile and that stronger growth must be supported by other indicators. We forecast GDP will increase by an average of 4.0% in Below we highlight the most important reasons for this assessment and show how this development differs slightly from that proposed by other forecasters (that remain with a less bullish growth outlook). INTERNATIONAL SLOWDOWN MAIN DOWNSIDE RISK. The more severe slowdown in the international business cycle is the main downside risk to the outlook for Swedish growth, and not only because it would result in significantly weaker conditions for exports and manufacturing investments. Further, international downturns almost always result in setbacks to private consumption as households become more cautious towards increased spending. INDUSTRIAL RECOVERY BACK ON TRACK. Though the outlook for global growth is a downside risk, we believe the recovery in manufacturing PMIs over the last month supports our view that growth concerns at the beginning of the year were exaggerated. There are also signs of improving manufacturing PMI in Sweden and a steady upward trend for goods exports that has extended into Q1, implying that Swedish industry is in better shape than many had feared. Indeed, over the last three quarters, goods exports have been stronger than at any time since A slowdown in service exports from a high level means that total exports are predicted to grow by around 6% in 2016, approximately the same rate as last year. However, manufacturing investments are predicted to be unchanged this year, after being relatively strong in PUBLIC SPENDING EXPECTED TO ADD 1.3%-POINTS TO 2016 GDP GROWTH. Increased public spending driven by immigration has only just begun to affect GDP. Costs for providing housing, health care and education for the 160,000 immigrants that arrived in Sweden last year are forecast to increase government expenditure by around SEK 40bn. Indeed, we believe this estimate may be too low, as it excludes various additional municipal costs and investments. Higher government spending of SEK 40bn corresponds to 1% of GDP and almost 4% of total public consumption, although some crowding out effects of other consumption is probable. We predict public consumption will grow by 3.5% this year. Furthermore, the need for public services for immigrants means that public investment in schools, hospitals and housing will have to increase. We expect public fixed investments to rise by 10% in 2016 (compared to 1.5% in 2015). The contribution to GDP from public consumption and investments is 1.3%-points in 2016, its fastest growth since the 1980s. Currently, the influx of immigrants to Sweden is well below even the immigration authorities lowest scenario for Disregarding whether this trend continues or not we expect a relatively small effect on consumption and investment forecasts. Most predicted spending in 2016 is driven by immigrants that arrived last year or earlier. 9

10 HOUSEHOLD SECTOR STILL APPEARS FIRM. Fuelled by strong real wage growth, rising employment and partly strong population growth, private consumption is gradually accelerating with short term indicators suggesting the trend is continuing into Car registration remains at cyclical highs, while consumer confidence has recovered as households take a more positive view of their own financial outlook. We see strong evidence that private consumption will continue to grow rapidly in RESIDENTIAL INVESTMENT WILL CONTINUE TO RISE. Residential investment contributed 0.7%-points annually to GDP in 2014 and In 2015, although housing starts increased to their highest levels since the early 1990s, the upturn lost some momentum at the end of last year. We forecast the contribution to GDP from residential investment will decline to %-points in A temporary upturn for own home repairs ahead of the reduced tax deduction at the end of 2015 will also contribute to this development. Despite this slowdown the combined contribution to growth from the public sector and residential investment will be %- points larger than last year. 10

11 Risk of taper tantrum a la Riksbank We expect the Riksbank to remain on hold at its monetary policy meeting on April 21 and to keep both the repo rate and the rate path unchanged. We do not forecast any new QE measures although we predict that the board will continue to re-invest maturing bonds and coupons. Also, we believe it will indicate in its press release that the threshold for introducing further monetary policy easing is being raised. Given market speculation that the board will introduce more QE in April we see a risk of taper tantrum in Sweden. RIKSBANK ON HOLD IN APRIL. The Swedish central bank is expected to keep monetary policy unchanged in its Monetary Policy Report, which is due to be presented on April 21. Most likely the board will also retain the 3bps easing bias in the repo rate path and reiterate its readiness to expand policy further if needed. Although the more formal components of its forecast are expected to remain unchanged we believe the board may make various small tentative changes to its language in the press release that will indicate that the bar for further expansion has been raised. This would mean that the board majority had taken a step towards adopting the views previously expressed by Cecilia Skingsley, who in the latest minutes indicated that the Riksbank should going forward hold the trigger, tolerate inflation deviating from the forecast and allow a longer period for inflation to be brought back on target. We expect the forecast for CPIF in 2017 to be lowered, to indicate that this strategic shift has begun. RISK OF TAPER TANTRUM A LA RIKSBANK. If the Riksbank decides not to extend its QE program, we see a risk of a similar, albeit smaller market reaction compared to the US taper tantrum in If we are correct the yield curve will become steeper and the krona stronger after the rate decision. To mitigate the effects of tighter financial conditions, we expect the board to continue to reinvest maturing bonds and coupons. Possibly it may also postpone the date of the first rate hike from mid-2017 according to the February report. RISK SCENARIO - EXTENDED QE PROGRAMME. With the current QE program set to expire at the end of June, i.e. one week before the monetary policy meeting at the beginning of July, the board may argue that it must be pre-emptive and announce a prolonged program at the April meeting to contain risks of a stronger krona. Recent strengthening by the SEK may potentially trigger more QE if it continues. Currently, the KIX currency index is 1% stronger than the Riksbank s own estimate for Q2. Inflation-linked bonds are likely to be included in any expansion of the QE program. CURRENCY INTERVENTIONS MORE DISTANT NOW. The message from the G20 summit in Shanghai was that member states would refrain from using the exchange rate as a tool to achieve inflation targets. This underlines how difficult it would be for the Riksbank to use direct currency interventions. Comments from Mario Draghi at the ECB press conference in March that exchange rate movements were a consequence of, but not a target for ECB policy further emphasise the political difficulties the Riksbank would face if it were to try to weaken the currency. Comments from different board members indicate that interventions currently are off the agenda. We think it will require a significantly stronger SEK before they return. Probably the board will continue to mention the possibility for interventions in its press release to keep market expectations alive, at least to some extent. BOARD FOCUSING MORE ON CAPACITY UTILISATION. In the slightly longer term we reiterate our scenario presented in the two latest editions of Nordic Outlook that rising capacity utilisation and inflation expectations mean that the Riksbank will refocus its attention during the second half of this year, to consider the best time to start hiking rates. A very important indicator for the development of this process will be measures of capacity utilisation, with the central bank s own capacity indicator 11

12 based on a large number of labour shortage indicators being the most important. The indicator is currently close to the historical average, but still below levels where wage and inflation pressures normally start to build. The labour shortage in the public sector is increasing faster, and is also at significantly higher levels than historically. This implies that wages could start rising at an earlier stage than in previous business cycles, while on the other hand an environment characterized by very low international price pressures means that despite this inflation may face difficulties rising. We think that it will take significantly longer for the cyclical upturn in inflation forecast by the Riksbank to occur. The labour market has developed largely in line with the Riksbank s expectations, with unemployment declining gradually. Further, employment growth is firm and accelerating slightly. We expect the unemployment forecast to be largely unchanged. NO MORE EXPANSIONARY MEASURES IN THIS CYCLE. In our view, it will require the recovery to be derailed or a significantly stronger krona for the Riksbank to expand its monetary policy further. We predict a first rate hike in February next year, though expansionary ECB policy and the risk of more gradual tightening from the Fed imply a risk that rate hikes will be delayed also in Sweden. STRONGER GROWTH AND SLIGHTLY HIGHER NEAR TERM INFLATION. Significantly stronger growth during Q4 implies that the Riksbank will raise its 2016 GDP forecast slightly from the 3.5% increase predicted in the February report. Lower sentiment indicators and probable downward adjustments to the international growth forecast (albeit small) will work in the opposite direction. We think that public consumption and investments will compensate for a slightly weaker manufacturing sector this year (read more here). We expect the Riksbank to interpret cautiously extremely strong Q4 growth. Average growth in 2015 was 0.5%- points higher than the Riksbank s own forecast. HIGHER INFLATION AT THE BEGINNING OF CPIF inflation at the beginning of the year has been %- points above the Riksbank s estimate with CPIF ex energy 0.2% higher than expected. Taken together with a slightly increased petrol price this implies an upward revision to the 2016 inflation forecast. Looking further ahead, the central bank s inflation forecast looks over optimistic given low international price pressures and signs that upward pressure from earlier krona weakness is easing. We believe the Riksbank will predict a more gradual increase in inflation in Still, a marked upturn for service inflation should make the Swedish central bank more confident that inflation is moving back towards its target. Inflation expectations have in most instances increased with the steady upward trend in break-even inflation, and have also moved higher according to the Prospera survey. However, they have risen most in the near term but less on the 5y horizon. Also, all expectations are below the 2% target. Still, higher inflation projections are likely to come as a big relief for the board. 12

13 What we expect the Riksbank to forecast on April 21 Riksbank Riksbank expected forecast in Feb CPI CPI CPI CPIF CPIF CPIF GDP GDP GDP Unemploym Unemploym Unemploym Employment Employment Employment Wages Wages Wages KIX KIX KIX Int. GDP Int. GDP Int. GDP

14 Service inflation good news but more needed Swedish service inflation accelerated to 2% y/y in January this year from less than 1% y/y at the end of Partly, the upturn signals that domestic inflation pressures are starting to build, which will be warmly welcomed by the Riksbank. Still, some of the upturn is driven by temporary factors likely to be reversed at the beginning of next year. Also, service inflation must increase significantly before total CPIF inflation is likely to revert to 2% in an environment characterised by low international price pressure and decreasing support from a weak exchange rate. The tighter labour market implies that inflation pressures should continue to rise in the medium term. HIGHER DOMESTIC INFLATION, BUT STILL NOT ENOUGH TO REACH TARGET. January CPI most likely pleased the Riksbank. Not only did it exceed both their and consensus estimates, it showed an upturn also driven by more rapidly rising domestic service prices, which were sharply higher. Service inflation accelerated from a low 0.8% y/y in December last year to 2.0% y/y in January. Despite easing slightly to 1.7% in February service inflation stands at its highest level since Although the acceleration is also driven by various likely temporary components, the prolonged historic period of very low service inflation appears to have ended. for zero boundaries for deposit rates to households. The contribution from other services, which probably best reflects the underlying trend, increased to 0.7%-points from 0.3%-points in December. Service inflation over the last 3 months 2015-Dec 2016-Jan 2016-Dec Total There of contribution from Rents Tax on own home repairs International travels Health care Banking services Congestion fees Other Assessing the underlying trend for services is not straight forward but excluding the impact from own home repairs is obvious. Currently, service inflation (excluding own home repairs) is slightly below the historical average. One reason for this situation is the very low rent increases secured this year, with rents set to rise by only 0.7%, due directly to lower mortgage rates driven by Riksbank rate cuts. A rent increase in line with the historical average (2.1%) would add 0.4%-points to service inflation. There are therefore several temporary factors affecting services also on the downside. We expect rents to rise faster in 2017 neutralising most of the base effects from own home repairs. To what extent the uptick in administrative prices and bank services will prove to be temporary is more difficult to judge, but pressure on municipalities and high demand for dental services indicate that these prices will continue to increase faster than in 2013 and The table below shows the breakdown of the upturn for service inflation in February compared to December. Tax hikes on own home repairs added 0.3%-points to service inflation, while a number of administrative prices such as congestion fees and health care added a further 0.3%- points. Furthermore, banking fees increased more than normal, probably due to banks seeking to compensate 14 WAGES STILL SUBDUED. So far, wages do not signal any large upward pressure on service prices, which supports the conclusion that several components of the upturn in service inflation at the beginning of this year will be temporary. The current wage round looks likely to result in a negotiated wage settlement largely in line with the rates contained in the three-year agreement in A gradually tightening labour market implies that wage drift (wage increases above negotiated wages) this year will be slightly higher. However, apart from the public sector labour shortage indicators are still moderate.

15 However, with 1-year agreements having been concluded this year the next wage round will take place in 2017 when capacity will be significantly greater. Prospects for higher wages in 2017 have increased. MORE SUSTAINABLE UPTURN IN SERVICE INFLATION FIRST IN THE MEDIUM TERM. We conclude that service inflation will remain around current levels this year, but fall back at the beginning of 2017 due to base effects from the temporary factors at work this January. However, an increasingly tighter labour market is expected to lead to a sharper acceleration in wage inflation in 2017, which will make service inflation resume its gradual upward trend. Nevertheless, the upturn we predict is, however, insufficient to bring inflation back to target before the end of 2017, especially in an environment characterised by low international prices and decreasing upward pressure from a (previously) weak exchange rate. Assuming capacity utilisation continues its upward trend in line with our expectations inflation pressures are likely to build even further, though a more pronounced upturn in inflation will probably not occur until after the end of the current forecast horizon. Also, it will most likely need support from higher international price pressures. 15

16 Sweden: Cautious Spring Budget Bill, improving government finances but risks ahead Swedish public finances have surprised on the upside. Positive forces dominate as revenue increases exceed higher spending by both central and local government, mainly due to greater migration. However, the longer term outlook is of greater concern with strong finances being driven by cyclical factors, as the GDP gap is closing rapidly. Because government finances are highly sensitive to the business cycle, changed assumptions on growth or interest rates could significantly alter the outlook going forward. While we expect a cautious Spring Fiscal Policy Bill on April 13, if public finance data remain strong, the Autumn Budget Bill for 2017 may be more expansionary. CAUTIOUS SPRING FISCAL POLICY BILL, BUT GOVERNMENT EAGER TO DO MORE LATER THIS YEAR. In the near term, the government faces large upward pressure in several expenditure areas that threatens the central government expenditure ceiling. Difficulties in keeping spending under the set limit are primarily associated with Additional reforms are expected to affect mainly the revenue side while spending reforms will take place next year. In particular, the government has promised to provide local governments with an additional SEK 10bn in Recent better-than-expected data and more positive views from, for example, the National Financial Management Authority indicate there is more room for manoeuvre under the expenditure ceiling than had been previously thought. Pressure on the government to spend more will come mainly from the Left party. The Finance Minister appears to be focused on keeping spending under the expenditure ceiling, though the government is also eager to move from reacting to migration and other issues to instead pursuing its own initiatives. All in all, the Spring Fiscal Policy Bill (due on April 13) is expected to include limited new policy initiatives. However, it will be interesting to view the government s latest assessments of spending pressure on migration and sick leave, as well as indications regarding what it will focus on in its Autumn Budget Bill, in addition to extra funding for local governments. Also newsworthy will be government plans to ease pressure on the housing market; discussions with opposition parties have so far been disappointing. Overall, fiscal policy is slightly expansionary in On the one hand, expenditure is increasing and 16 government consumption and transfers boost growth, substantially driven by migration-related spending and investments. GOVERNMENT FINANCES HIGHLY SENSITIVE TO THE BUSINESS CYCLE. With high marginal taxation and an extensive benefit system the central- and general government balance fluctuates significantly with the business cycle. The EU and OECD estimate the government s finance/gdp elasticity at between 0.6 and 0.7, i.e. 1 pp in higher (lower) growth would improve (lower) the government balance by % of GDP. Our estimates imply that even if cyclicality has probably decreased over the past decade due to tax-cuts and lower social transfers, elasticity in the present cycle could in fact be even higher as actual elasticity varies depending on which parts of the economy fluctuate and how heavily taxed they are. The effect is less for increased manufacturing exports and higher for construction and consumption. This is because these sectors generate significantly more tax revenues, but also because the impact on employment is significantly larger. The cyclical nature of government finances was clarified on March 31; Statistics Sweden reported that government finances in 2015 were in balance for the first time since 2010 and up SEK 61bn from -1.6% of GDP in Most forecasters, including us, estimated only a few months ago that the deficit would be between

17 0.5 and 1.0% of GDP. Also, the central government borrowing requirement has proved better than expected for some time. Over the last 12 months, the central government budget has even registered a small surplus. DESPITE MIGRATION SPENDING, MANY FACTORS ARE IMPROVING GOVERNMENT FINANCE. Both permanent and temporary factors are contributing to an improvement in government finances, which we expect will remain strong in the short term. Nevertheless there are also problems ahead. At present, growth is partly boosted by imbalances in the economy; migration fuels government consumption, lack of housing fuels construction. Structural misjudgements have caused and will cause problems that exacerbate risks. While migration is much in focus, upward pressure on government expenditure is more broadly based and includes, for example, defence, police, grants to municipalities and labour market issues. Local governments are under considerable pressure; increasing difficulties in finding qualified staff may drive up wages while migrants impose strains on social services, hospitals and education. Demands for more central government grants will remain high. Recent favourable government finance outcomes and the eventual borrowing requirement hide various risks going forward. Current examples of positive factors/risks ahead right now are as follows: 1) Interest on government debt is extremely low. The latest estimate by the National Debt Office (NDO) is that interest expenditure on government debt will be near zero in According to the definition used in national accounts, the interest cost on central government debt is down to 0.5% of GDP from 3.6% in 2000 and over 5% of GDP during the mid-1990s; 2) Taxes on capital gains have risen sharply. This item is often one of the main drivers for large fluctuations in the government balance; 3) Tax revenues generally are increasing rapidly in line with the fast growing economy. In particular, those sections of the economy that have been generating high tax revenues such as employment, consumption and construction have been important drivers; and 4) Unemployment related expenditure is falling. RISKS AHEAD TO WHAT EXTENT WILL A DIFFERENT MACRO-SCENARIO AFFECT GOVERNMENT FINANCES? What could happen to government finances if the future differs significantly compared to present forecasts? Higher/lower growth: Using OECD or EU government finance/growth elasticity only, 1pp lower growth would have a negative effect of approximately SEK 25-30bn on the government balance. This figure refers to the entire government sector, so the impact on the central government balance depends on the composition of the slowdown. The unconsolidated central government expenditure/revenue share of the general government sector is approximately 50%. As the central government is exposed to most of the effect of the cyclical forces acting on government finances, the approximate impact on central government may amount 17 to up to up to SEK 20bn (i.e. 1pp lower GDP growth could worsen central government finances by SEK 20bn or 0.5% of GDP). Higher/lower interest rates: Government debt is around SEK 1,350bn an increase in interest rates of 1pp would therefore raise interest expenditure by SEK 13-14bn ( % of GDP). As the average maturity of central government debt is approximately five years, it takes time for increased interest rates to be reflected in higher interest expenditure. The National Debt Office rule of thumb is that an interest rate rise of 1pp boosts expenditure by SEK 5bn (SEK 2bn for a 1pp higher international interest rate), which represents a higher feed-through effect than in our top-down approach. Migration spending: Since the steep rise in the number of asylum seekers last fall, spending forecasts have risen sharply. Compared to the sum appropriated in the Budget Bill for 2016, the migration agency now asks for around SEK 30-35bn more for this year. Although it seems likely that the number of arrivals has fallen (Jan- March this year saw the lowest number of arrivals since 2012), spending will be high for a prolonged period. Pressure to spend will continue to increase firstly for the migration agency but later also for municipalities and other social insurance systems. It seems like the upside risk has fallen with arrivals now lower (though forecasts are very uncertain), but major questions remain regarding the length of time it will take before a migrant actually lands a job. It is difficult to assess the impact on public finances of the full effects of migration, although given the costs associated with large flows, and major budgetary revisions during the past six months, clearly this area may have both positive and negative effects for the budget going forward Asylum seekers (monthly) Illness and early retirement: The appropriation for sickness leave (according to the Budget Bill for 2016) will increase from SEK 32bn in 2014 to SEK 51bn in During the same period, the appropriation for early retirement is stable. Overall un-health data (average paid out days in total social insurance) is stated despite the expenditure increase for sickness leave being fairly stable or just increasing slightly. As regards the NIER s longer term forecasts, expenditure on un-health as a

18 percentage of GDP is expected to be stable to So overall we see a disturbing upward trend in sickness leave, although so far a decreasing number of persons in early retirement has worked in the opposite direction. In 2003, total costs for sickness (sickness and early retirement) peaked at 4.0% of GDP and stands currently at approximately 2% of GDP. This highlights the fact that costs can increase rapidly if politicians fail to control sickness expenditure. TIME FOR NEGATIVE SHOCKS IS PROBABLY SOME WAY OFF STILL. While our calculations above are rough, they highlight that rapidly occurring changes can have significant effects on government finances. Compared to a baseline scenario, a 2pp higher interest rate would worsen the central government balance corresponding to 0.7% of GDP deficit, or growth falling by 2pp could worsen government finances by more than 1% of GDP. Swedish government finances are in good shape. However, the main reasons that the balance appears sound at present are the same as those that might cause it to deteriorate rapidly in a downturn. In our view, good news will outweigh bad for some time yet. Nevertheless, we believe the present positive development in government revenues will not last forever. GDP and revenue growth will slow going forward though forecasts for look strong. Effect on central government borrowing requirement of higher interest rates or lower GDP Three scenarios 1, 2 and 3% change compared to baseline, effect in SEK billion 1% 2% 3% Higher interest rate Lower GDP Note: The effects represent broad assumptions applying GDP/government finance elasticities and the long-term effect on government debt of a higher interest rate. The columns show the effect on central government borrowing requirement if interest or GDP would be 1% (2%, 3%) higher 18

19 To Brexit or not to Brexit? Sweden more politically affected than economically by UK decision The Brexit issue affects us all, whether we want it to or not. While politically, economically and financially the issue mainly concerns the UK, it will impact Sweden. However, the economic effects of the country s decision are likely to be fairly minor, especially in the event of a Yes vote. For Sweden, a decision by the UK to leave the EU will increase uncertainty, lower growth and push both yields and the SEK lower. YES, NO OR NYES? At present polls suggest the UK will vote Yes to continued EU membership in the upcoming referendum, though there are still a large number of undecided voters. We outline various alternatives in this note that include Yes, No and Nyes votes. A Nyes outcome involves a small No majority (or soft No) that results in further negotiations with the EU, which are then followed by a second referendum where we would expect a Yes vote. We regard a Yes vote as 65% probable and a No vote as 10% likely. We see a 25% likelihood of a Nyes vote. A YES WOULD HAVE THE SMALLEST CONSEQUENCES FOR SWEDEN and the forecast associated with that alternative are in line with our main scenario. As we consider its overall effects independently of the outcome of the referendum, we regard them as more political than economic, mainly because any new opt-outs negotiated by the UK would affect Sweden s situation. For example, might Sweden follow suit and demand similar opt-outs concerning membership of an ever closer union? Politically, this is an important issue as opt-outs even more clearly than before formalise in writing that the EU is not following a two-speed road forward, but instead a two-track one. Also, in an EU without the UK, Sweden would have a harder time finding a role. It would then be forced to make a clearer choice regarding its role in the EU and its attitude towards deeper cooperation in an increasingly federalist Europe. In the event of a hard No, Brussels and core EU countries may exert pressure on Sweden to adopt the euro because such a decisive rejection of the union by the UK might encourage the EU to pursue further and faster integration as the way forward. If faced with the choice of accepting such a policy or leaving the EU, we would expect Sweden to leave the union. LIMITED SHORT TERM ECONOMIC IMPACT. In the short term, the economic impact on Sweden is unlikely to be large. However, especially in the event of a hard No vote, Sweden would be affected by any European trade disruptions. Also, hopes that Stockholm might strengthen its competitive position as a financial centre if the UK becomes more isolated seem far-fetched. THREE SCENARIOS IN FIGURES. Compared to our present main scenario for global and Swedish growth and financial markets, a No vote would have the largest effects. Rejection of continued EU membership would dampen GDP growth though the impact would probably be limited in 2016 and The main consequence would probably be greater uncertainty, for example concerning trade legislation and financial regulations, holding back exports and investments in Europe. The UK would be hardest hit, mainly in terms of investments as companies may prefer to invest inside the EU rather than in an isolated island country. Partly, a weaker GBP would offset these effects on the export side. The accumulated effect on UK GDP of a Hard No scenario would slightly exceed 1% in 2016 and In the euro zone and Sweden the corresponding figures would be nearly 0.5%. In the long run, potential growth could suffer somewhat due to less dynamic underlying economic conditions in Europe. Table Effects on Swedish growth Yes Soft No Hard No LOWER SWEDISH YIELDS IN NO-VOTE SCENARIO. Long-term UK yields would increase slightly if the UK voted to leave the EU, due to an increased risk premium, although the Bank of England will stand prepared to provide further stimulus if necessary. Conversely, German and Swedish government bond yields may be driven somewhat lower in the event of a No vote, being viewed as safe haven/ alternative investments. The Bank of England (BoE) may delay key interest rate hikes if a No vote weakens economic conditions => a steepening of the UK yield curve. Table Swedish 10Y yield Sep 2016 End 2017 Yes Soft No Hard No SEK TO SUFFER DUE TO INCREASED UNCERTAINTY. In the event of a No vote, the initial reaction is likely to affect the GBP quite considerably. The EUR would probably also suffer against the USD, reflecting general uncertainties regarding future economic and political cooperation in Europe. But since a soft No vote would probably trigger political negotiations that could open the way to a second referendum, the reaction would likely be reversed more rapidly. In a hard No scenario, the GBP would remain weak. The SEK would suffer against the EUR in a more uncertain environment following a No vote. 19

20 Table SEK to suffer in more uncertain environment Sep 2016 End 2017 Yes Soft No Hard No

21 Position for a flatter 5y/10y SEK vs. EUR curve Expecting the Riksbank to leave its policy unchanged at its April meeting and with additional easing measures looking increasingly unlikely, markets are set to refocus on future rate hikes this autumn. With the 5y/10y SEK swap curve extremely steep and the belly of the curve predicted to underperform during H2, we recommend positioning for a flatter curve, preferably vs. the EUR curve. We also see value in receiving 5y5y SEK vs. EUR IRS. We find short dated Swedish inflation-linked bonds very cheap vs. nominal bonds and recommend BEI longs. Finally, as a downside hedge we recommend selling Dec16 RIBA futures at -0.51%. APRIL 21 MEETING. We expect the Riksbank to maintain the monetary policy unchanged (retaining a 3bps easing bias while avoiding presenting new QE measures) on April 21. Through its communication, we believe the central bank will convince markets that it stands ready to ease policy even further if deemed necessary, even between ordinary monetary policy meetings. Such an outcome would likely lift the 2y yield by 1-2bps. With markets divided whether new QE measures will be presented in April (around 40% for government bonds, 20% for IL bonds, close to 0% for covered bonds according to our investor survey), the 10y yield will most likely increase by approximately 2bps (relative to Germany) if new QE measures are announced. Most likely scenarios for April 21 (our subjective probabilities) Main scenario: 65% probability Dovish risk scenario: 30% probability Very dovish risk scenario: 5% probability Unchanged rates, easing bias remains, no changes to QE programme Unchanged rates, easing bias remains, SEK 50bn new QE 10bps repo rate cut, easing bias remains, SEK 50bn new QE Dovish risk scenario. There is also a relatively high likelihood that the Riksbank will announce new bond purchases on April 21. If it does, we expect the 2y yield to decline 0-2bps and the 10y yield to decline slightly (5bps). Very dovish risk scenario. A combination of a rate cut (10bps) and a QE announcement would represent a major dovish surprise for markets, depressing the 2y 21 yield by ~10bps and lowering the 10y yield by around 7-8bps. SELL DEC16 RIBA FUTURE. As a tactical position in rates ahead of the April meeting, we recommend selling the Dec16 RIBA future at -0.51%. At its maturity at IMM Dec16, the contract will fix at -0.50% if the repo rate remains at its current level and will generate profits if the repo rate is cut before the contract matures. While our main scenario does not include another repo rate reduction, the risk/reward involved in selling the Dec16 RIBA is very attractive as a rate cut may be made later this year if the macroeconomic outlook deteriorates or the ECB announces more stimulus measures. The position also acts as a hedge ahead of the UK referendum (Jun 23), where a No vote would create increased uncertainty and financial market turbulence, potentially triggering a dovish response from central banks including the ECB and Riksbank. MARKET TO REFOCUS ON RATE HIKES DURING THE AUTMN BELLY TO UNDER-PERFORM. As Swedish short rates are lower than their EUR counterparts, and because of the sharp widening of SEK vs. EUR spreads further out on the curve, the 2y/10y SEK swap curve is historically steep. Even more so, the 5y/10y curve is at very extreme levels, not far from all-time highs of around 100bps. Over the next 6-12 months, we expect the 5y/10y SEK swap curve to flatten both outright and relative to the EUR. This is because we believe the Riksbank will initiate rate hikes well before the ECB, with markets likely to refocus on upcoming increases already sometime during H As a result, SEK vs. EUR rate spreads will widen especially in the 2-5y segment but less so in the 10y segment, resulting in relative flattening of the SEK 5y/10y curve vs. EUR. Convexity of the 2y/5y/10y SEK IRS curve is near record lows and is set to increase when the 5y rate starts to rise. We expect this to occur when markets refocus on future Riksbank rate hikes, helping flatten the 5y/10y curve. From an historical perspective, current 5y/10y SEK swap curve steepness is extremely high relative to the convexity.