Debt and equity capital markets

Using near-term forward spread, namely the spread between 6-quarters ahead forward rate and the current yield on the same T-bill instead of long-term bonds’ yield difference such as T-10 years and T-2 years, could stochastically dominate the usual spread. Whereas the T-10 and T-2 spread is somewhat indicative of a recession – because the spread signals near-inversion of the yield curve - the near-term forward spread doesn’t trend downwards, claimed a recent Fed study. In as much as the near-term forward spread is only the mirror image of market expectations, it only conveys what market players think, and doesn’t spread a data-generating process of its own. So, there was no high recession probability signaled by the near-term spread, and if it is a better measure than the usual yield spread because it tracks market expectations more closely, the odds of a monetary policy change were rather low. I therefore note that the idea that the Fed is nearing its tightening cycle was quite premature. Until recently, that is.

Debt and equity capital markets may not be equally sufficient statistics as predictors of an incoming recession but they convey quite useful information regarding the real sector in their own ways. Of course, they are also important per se. Samuelson’s joke about it is well known, saying that stock markets had predicted 9 out of the recent 5 recessions. Be that as it may, recent stock market sell-offs could indeed have left an indelible dent in the minds of Fed policy makers. Hence, the world hasn’t turned upside down perhaps, but the vision has drastically changed in a single month after the $14 billion global sell-off, of which one third came about in the U.S. stock markets. On the other hand, debt markets are a better indicator of what may come next. Yield curve inversions is one such example. The loan market is another, provided of course that it isn’t always a lagged indicator. Let’s see.

W ll there be pol cy rate cuts somet me n 2019?

Yes. Or at least it is quite likely if things go well. In general, people resort to simple Taylor Rule computations in order to assess which central bank is most ‘behind the curve’. They can be tricky though by being deceptively simple. Suppose the potential GDP growth rate is still 5 percent. Suppose the effective inflation target is (still?) 5 percent. Consider a fifty-fifty growth/inflation Taylor coefficient set. What would you expect for 2019? It would depend on your target interest rate. If we assume that 6 comes from the coefficients set, assuming 17 percent expected inflation for 2019 – quite close to the current CBRT poll that stands at 16.46 percent now - you would come up with target interest + 11 percent. Assume the normal target is inflation + 6 this time around. So, you might come up with a 23 percent policy rate. The most you can get is 28 percent. Actually, the one-week combined repo rate is at 27.11 percent. The secondary market benchmark bond rate premium is at a discount, which doesn’t look very stable. Still, this is the highest it could get in our opinion and the discount is at its deepest. What is the effective current rate? 24 percent. At most the ‘behind the curve’ argument, from a purist standpoint, hints at a 400 basis points delay. Correct? No. We would rather think not, and set it at 23 percent - hence 100 basis points lower than the current funding rate - because 2019 growth is more likely not to be only below potential but very low. Hence, from the coefficients set there will come not 6 but 6 plus the growth coefficient, at least minus 2. So not 6 but 4, overall. That leaves us at 21 percent to 26 percent, the median of which is 23.5 percent, just about the current funding rate. Yet it is quite fragile a balance. So, better not toy with the CBRT rates. This is why the CBRT’s recent stance was widely approved.

Having said that, however, it is obvious that if inflation falls short of the current 2019 year-end expectations, the same Taylor Rule – with the same coefficients - would yield a lower target interest rate. Should expected inflation fall to 13 percent, for instance, instead of the consensus 17 percent, this would result in a slack of 400 basis points. Furthermore, this development could be a by-product of a prolonged slump. In that case, because GDP growth also will be less than expected, and additional interest rate slack will come from there because Taylor Rules generally attach weights to both inflation and growth, lower inflation and lower growth imply two avenues for rate cuts. I would then opt for at least 500 basis points sequential cuts in H2 2019.

Yet there is a caveat. The ex post real rate of interest and the ex ante real rate are depicted in the relevant graphic. Historically, there

has been a huge discrepancy, meaning that either inflation expectations have been systematically wrong or there has been a bias in favor of inflationary surprises. This tendency is clear especially after mid-2016. Between April 2013 – right before the ‘taper tantrum’ - and mid-2016 the distance between the expected and realized real rates wasn’t huge. After that, it varied a lot. Today’s 12-months ahead (expected) real rate is 3.68 percent but 12-months backward real rate is negative now. So, inflation has to fall beyond consensus and it has to fall trend-wise, for good.

Can banks del ver?

Let’s not be preoccupied with what has already been done. Bygones are bygones, and credit growth is almost dead now. It has to be refueled if growth is to come back in H2. Can it be revived? It depends, and not only on syndications. The impact of the Fed interest rate and liquidity management policy moves on EMs looks asymmetric in the sense that tightening has more of a marked adverse effect on EM capital inflows than loosening. However, a recent CBRT study implies, the share of an EM country is a better measure of its vulnerabilities. The flow of funds can decrease overall but a specific country can manage to attract a larger share out of a shallow pool. The Fed funds futures pose as a natural candidate with which to estimate the impact of the Fed’s future policies. As the expectations channel appears to be the main spill-over channel, all expectations-related phenomena – politics, business conditions, rule of law, Central Bank independence etc. - turn out to be crucial in securing a satisfactory flow of funds in a world of tightening. The policy rate is perhaps key but this metric is not a sufficient statistic by itself. Now consider the curvature of the TRY yield curve and loan growth. If it is correct that the level is a long term factor associated with inflation and the slope co-varies with the business cycle, then it may be that the curvature of the yield curve is associated with financial conditions. It is related to the cycle also. Domestic financial shocks matter more than they did in the past compared to inflation shocks, and weigh on the yield curve’s curvature, which in turn may have a bearing on its slope.

A recent CBRT study estimates the swap yield curve through Nelson-Siegel. This we also find very promising because the Turkish banking sector has not only relied on ‘on balance-sheet’ external funding but also off-balance sheet contracts – mostly cross-currency swaps – which reached $50 billion. The fitted zero-coupon swap rates and forward implied (exchange) rates largely coincide, meaning it is possible at least to make some inferences based on the swap curve. Now this can be important because swaps are a crucial vehicle through which banks lend long term, as in mortgages. On this front, the development of the USD/TRY swap curve since mid-August is promising as underlined by the CBRT a few months ago.

The argument n favor of a Turk sh recovery n H2 2019 – w thout the IMF

MSCI Turkey premium over EMs has fallen from minus 67 to minus 54 within a span of four months. Actually, it has been there before August, but the beginning of 2018 witnessed 35 basis points only. This is akin to the trajectory of the exchange rate. All in all, there is a hit but, both in terms of mean and volatility, the new situation is better than mid-2018, a consolation perhaps. On the other hand, it might signal a gradual return to normalcy. There are two scenarios now. First, the automatic normalization helped by the Fed’s new stance, ECB’s standby position, China’s slow down and oil prices hovering just above $50 gives us a new vision. This is, in the innermost core of the appearances, possibly the downturn of the global tide because capitalism is cyclical. If this is so, then global interest rate hikes may have come to an end, and even if the U.S. economy doesn’t enter a recessionary phase any time soon, interest rates there may at least stay flat for a long time to come, if not fall. In this scenario, there can be two versions. Primo, in H2 2019 the Turkish economy will recover if not rebound, at least on account of favorable base effects, and since foreign funding won’t be any more expensive, credit channels will reopen. Then, the whole of 2019 may post positive growth, around 1-1.5 percent. Secondo, even so, the H2 recovery will delay, and we may end up with negative growth by the end of 2019. Everybody agrees that Q4 2018, Q1 2019 and possibly Q2 2019 will showcase a recession. In this first general scenario, equity markets will perform, on account of quite low multiples for banks, expected good performance for exporters, and even for growth investors there may be found some strong balance sheet/no FX-debt firms. The stock exchange as a whole provides an upside then. Obviously, because inflation will also fall, probably somewhere between 13-17 percent by the end of 2019 barring any exogenous shock and depending on the speed of the recovery, discount rates will also fall – as they are already falling now - which greatly affects company valuations.

The second scenario claims that the relative improvement in financial markets is only temporary and the malaise is so deep that Turkish corporates will warrant a large capital injection from the outside world. In this case, ‘muddling through’ without the IMF won’t deliver anything, and after the elections we will witness a standby of some sort.

I impute a higher weight to the first scenario, a conjecture I will develop in detail next week.