Presenting Mr. Crisis – Summa Cum Laude
“All that is gold does not glitter Not all those who wander are lost; The old that is strong does not wither, Deep roots are not reached by the frost, From the ashes a fire shall be woken, A light from the shadows shall spring.” The Fellowship of the Ring, J.R.R. Tolkien
Well, that is the end, folks! We have finally come to the anti-climax and first, all positive expectations and then all the calamity forecasts are borne out by events; they have become realities. Before the advent of a new era, let us state plainly the stylized facts of the Turkish economy once again.
Enters external balance first: One, the current account (CA) does not determine the level of the nominal exchange rate. Two, growth implies current account deficit. Three, the real exchange rate is useful and significant in any regression with the current account as a variable to be explained.
However, the strong relationship between the current account and the real rate of exchange weakens to the point of fading away if other explanatory variables are included in the model. In fact, binary relations are desultory more often than not and we should not take the coefficients generated by bivariate models too seriously. Now that the trade-weighted real effective exchange rate is at an almost all time low, exports don’t boom, which testifies to the rather subordinated role of the real exchange rate as a determinant of export performance. Also, the CA and the net errors and omissions item influence the exchange rate. But their coefficients easily lose significance when other variables are included in the regressions, i.e. interest rates and debt service.
Four, oil prices and the energy bill were extremely important in the last 22 years in terms of exogenously and inelastically constraining the CA deficit. And that is the first pillar of the anti-climax, of which more below.
There is a world pillar in the Finnish mythological epic Kalevala. It was painted by the Finnish artist, Akseli Gallen-Kalela, in the 19th century. The loss of the world pillar, or the foundation on which the olden world lies, causes a great trouble. As in the symbolic family of “rings” in the Nordic culture, the importance imputed to an outside object in which the talent of the age has been invested constitutes the core of the legend. The ring was an extremely important metaphor for die Jungkonservativen, from where a plethora of German nationalisms sprung and drove Hitler to power. I have reproduced the image in order to convey the message that the danger is all too big and that a sea-change in each and every economic, political and financial mindset should follow. The world pillar of the Washington Consensus was lost forever a decade ago. Now, all anchors - domestic and international - are being lost one by one.
As of 2000, the exchange rate was the most efficient nominal anchor and the pass-through coefficient was quite high. After almost two decades, it still is. It is true that TRY appreciation played a prominent role in bringing about the significant disinflation process we have witnessed in the first half of the decade after the 2001 crisis, although disinflation came to a halt in 2006. The two main years of steep disinflation were 2004 and 2005. Inflation has been trend-wise stable after 2006. It could have gone up to settle in the double-digits in 2008, but it didn’t and it seemed very likely that disinflation would follow the next year, due to recessionary dynamics, lack of demand, and rapidly falling oil prices. Nominal exchange rate realignment is an inflation propeller, yes. However, neither inertia nor the pass-through effects were as strong after 2008 as they were back in 2000. The pass-through coefficient had weakened by a good 10 percentage points in the three years after 2001, and weakened even further after that. Even if we do resort to a disaggregated approach – as one recent CBRT study does - and estimate that coefficient at a higher level than the traditional unrestricted VAR approach does, it is now around 15 percent for the CPI. It is fair to say that TRY appreciation triggered disinflation in the beginning but that it is no longer warranted as a disinflation driver. That the nominal rate of exchange does not jump discretely should have sufficed if disinflation were to continue. However, this didn’t happen and couldn’t happen because the only thing fundamental that has changed was a translation of fiscal deficit to the current account deficit and corporate FX debt. Turkey was journeying through the vicissitudes of fiscal deficit
in the 1990s; in the 2000s, it was those of the international trade deficit and FX-indebtedness.
The real rate of interest is one of the key variables more than ever and will remain so from now on. Because disallowing currency depreciation to take hold was the main target, interest rates ought not have to fallen in real terms throughout the whole decade, even after 2008, and especially after 2013. Nevertheless, now the ex ante real rate seems high, yet inflation is anchorless and any high ex ante real rate seems to provide merely a necessary anchor, not a sufficient condition. Hence, the market has recently pushed the sovereign 2-year yield up to 25 percent already.
It has often been said that philosophy, like Minerva, always comes late. Economic theory also delays in many instances. Retrospective analyses abound in scholarly journals. But this is not always so and it has been known since the late 1980s that balance sheet or net worth effects ought to be incorporated into the balance of payments’ models. Does the risk premium of a country have anything to do with its exchange rate regime? In other words, does switching to a flexible rate regime suffice to insulate the economy from the advent of excessive risk premia? Or, is it true that the risk premium is unrelated to the choice of exchange rate regime? In conventional accounts, expansionary monetary policy and depreciation of the currency are optimal responses to an adverse exogenous shock. However, if the economy has a large stock of FX-denominated debt, a weaker currency can exacerbate debt service difficulties and damage the balance sheets of domestic firms and banks. Hence, devaluations may be contractionary, a fact we have observed in the past. Thus, central banks do not like devaluations and not only for fear of inflationary pressure. The interaction of dollarized debt and net worth complicates the economy’s response to external shocks. It is likely that under financial vulnerability, a real depreciation raises the risk premium. Nonetheless, trying to contain real devaluations should depress domestic output, which is also detrimental for the risk premium. The interesting thing is that Turkey did not live through any home-made contraction after 2002, save 2008 of course, although it was possible to say that we had been in a regime of financial vulnerability, and the TRY had been overvalued throughout. Then, after 2008, the lira has depreciated more than inflation for a decade. It didn’t help because even exports depended on imports. This era has now also come to an end and we are either in plain recession or on the verge of it. True, the risk premium may be unrelated to the choice of the exchange rate regime, at least in the sense that there are two contradictory, offsetting effects. Even so, Turkey did not go by the book after 2002 until 2008, although growth has been below trend since 2007. And we may now be entering a regime of financial and economic weakness in full symmetry with the relative financial and economic strength episode we experienced after 2002 up to 2008. People claimed a decade ago that Turkey may have reached to a structurally stable path. By 2013, however, this claim had already proven to be an illusion. It is all built on debt.
How about the banking sector? NPLs have been declining since 2001, though part of it came through the Istanbul Approach under which some $5.5 billion had been restructured. More importantly, NPL provisioning had also improved significantly. Noncore asset disposals went through unaffected over the last fifteen years. Balance sheets are much less loaded with non-financial equity stakes and real estate. Volatility of interest rates implies that the stability of a bank’s earnings depends largely on non-interest income, particularly income from commissions and fees. Historically, with the high interest margin earned from investments in government securities, Turkish banks have subsidized other banking business by offering zero fee transactions to gain market share. This practice has been discontinued and fees and commissions have started to contribute to the bottom line. The weighted average (i.e., weighted with assets) commission was a mere 1.4 percent in 2003. This was in line with ratios in transition economies, from a low of 1.3 percent for state banks to 1.9 percent for privatized banks. From there, which is almost from scratch, the system underwent a drastic change. The bulk of commission income is from the consumer business, mainly from credit cards and asset management fees. Net interest margins (NIM) have been declining somewhat since last year. This trend would continue if all means of monetary easing were used, and asset growth ought to be the unique remedy here. This of course implies that shareholder’s equity should rise in tandem. This isn’t what is going to happen. There will be a full stop there also. The problem was in the risk aversion curvature once: banks might exacerbate economic woes through unnecessary credit rationing after 2008. Now, the problem simply is that funding is far too expensive to pass on to the consumer. Demand for credit is falling also. Still, all in all, balance sheets are much cleaner now and the system as a whole is a lot healthier. This is the official line of course, and now restructurings come to the surface, rendering underlying risks visible, of which more next week. Nonetheless, banks are the core of this economy and they should remain healthy and ready for any new beginning, as in 2010. This is the ultimate line of defense.