TURKEY:WHAT IF INVESTORS HEAD FOR THE DOOR?

UNCLAS SECTION 01 OF 05 ANKARA 002070 SIPDIS SENSITIVE TREASURY FOR INTERNATIONAL AFFAIRS - MMILLS AND CPLANTIER NSC FOR BRYZA AND MCKIBBEN E.O. 12958: N/A TAGS: EFIN TU SUBJECT: TURKEY:WHAT IF INVESTORS HEAD FOR THE DOOR? REF: A. ANKARA 1559 ¶B. ANKARA 1318 This cable was coordinated with Congen Istanbul. ¶1. (SBU) Summary: The acceleration of--and Turkey's dependence on--short-term portfolio investment flows has revived fears of what could happen if these investors head for the exits. If an exogenous shock is big enough, this "hot money" will indeed flee. Structural changes since 2001, however, mean that there is a very low probability of a full-blown financial crisis. Among the key structural changes are the floating exchange rate regime, an independent central bank under capable management, and a stronger banking sector. Turkish Treasury would be able to service its debt and there is little systemic risk to the banking sector though there could be an indirect impact on banks through corporate foreign exchange exposure. On the other hand, a precipitous fall in the exchange rate could provoke a slower, non-financial but ultimately politically-difficult disruption. End Summary. -------------- -------------- Surging Portfolio Flows Revive Fears of a Sudden Reversal -------------- -------------- ¶2. (SBU) As previously reported, the first quarter of 2005 has seen a surge of portfolio investment into Turkey, with inflows in January and February alone roughly 45% of inflows for all of 2004. Turkey not only remains dependent on these flows to finance its large current account deficit (reftels) but the inflows have caused the lira to appreciate in 2005, exacerbating its presumed overvaluation and widening the current account deficit. The March mini-sell-off brought a correction in the exchange rate but inflows could easily resume, as could the currency appreciation. This surge revived longstanding fears of what could happen if these investors suddenly rush to the exits en masse. This fear has been exacerbated by substantial anecdotal evidence that a portion of the increased flows stems from new classes of investors who did not take on Turkish risk--particularly in the domestic Turkish financial market--prior to the December 17 EU decision. These investors are more likely to be skittish in the event of a significant exogenous shock. -------------- Exogenous (or Endogenous) Shock -------------- ¶3. (SBU) In Turkey, there is an ample supply of potential shocks. Fed tightening leading to a global sell-off in emerging market debt; government-military or government-presidency tensions; splits in the governing party; earthquakes; rifts with the IMF, EU or U.S.; terrorist attacks; or spillover from instability in Iraq or the broader region are among the more obvious possibilities. The question of which of these would be the source of the shock is very difficult to predict. Instead, this cable focuses on what would happen in the event such a shock were significant enough cause the short-term portfolio investors to try to get their money out of Turkish markets. -------------- Low Probability of a Financial Crisis -------------- ¶3. (SBU) Given the structural changes since 2001, none of our contacts assign a significant probability to a crisis situation like occurred in 2000-1, even in the mass exit scenario described above. Since, unlike in 2001, there is no fixed exchange rate to defend, we define a "crisis" as a situation in which the Turkish Treasury has trouble servicing its debt or in which there are systemic problems in the banking sector that threaten to spill over into the real economy. There could, however, be significant disruption to the economy, mostly arising from a sudden, precipitous fall in the exchange rate as the investors dump their lira assets. -------------- The Lira Would Take a Dive -------------- ¶4. (SBU) The most disruptive element of the this scenario would be a sharp fall in the exchange rate, perhaps on the order of 20 or 30 percent. The lira is widely considered to be overvalued, driven up by the portfolio flows. The absence of these inflows, combined with a rush to purchase foreign exchange, would send the currency tumbling quickly. The Central Bank--recently seconded by a public statement by Minister Babacan--has repeatedly warned that it will not facilitate investors' attempts to exit the market by buying up lira. Though the Central Bank has justified its occasional interventions in the FX market by citing "excess volatility" in 2004 and 2005 it has only intervened in one direction: buying up foreign exchange when the lira appreciated too quickly. By intervening against the lira, the Central Bank builds up its foreign exchange reserves. These reserves have grown in recent months to $38 billion--more than adequate to meet any foreseeable foreign exchange requirement in the coming months. The Bank insists, quite credibly, that it will not use its precious reserves to break the fall of the lira. ¶5. (SBU) The absence of help from the Central Bank means that when fleeing investors look for foreign exchange in Turkey's thin market, it will be scarce, and the exchange rate could fall very far very quickly. One potentially mitigating factor on the swiftness of the lira's fall, however, is that even in a severe shock scenario some investors may opt to keep some of their positions rather than take large losses in thin Turkish markets. Several market-watchers have made the point to us that Turkish markets are so thin that losses can be sizable in a sell-off; yet Turkey has come back so many times that investors have learned the way to make money is to not panic in a correction. Consequently, even if most investors flee, others are likely to to hang on, reducing the size of the outflow. -------------- Banking System Less Vulnerable Now -------------- ¶6. (SBU) The banking sector as a whole is likely to withstand a sudden, deep fall in the exchange rate. First, the 2001 crisis knocked out the weaker banks, leaving most of the remaining banks either better-managed or attached to strong corporate groups. Having gone through the experience of 2001, both bank executives--and especially bank regulators--pay more attention to risk management. Under pressure from BRSA, banks have been building up their capital adequacy ratios and gradually reducing their exposure to related companies (related company lending was a major cause of the 2001 banking crisis). Banks' balance sheets were also helped by three years of favorable macroeconomic conditions, and a profitable period of capital gains from government securities during a period of rapidly falling interest rates. According to BRSA data, the total banking sector's owner's equity has increased 30% in lira terms (more in dollar terms, because of the lira appreciation) from year-end 2003 to year-end 2004. State-owned banks' owners equity matched the overall sector's growth rate of 30% as well. Several Turkish banks reported record profits in 2004 and January data suggest they continue to enjoy healthy profits--the sector's net income totaled nearly YTL 1.2 billion (about $900 million). ¶7. (SBU) With regard to the specific issue of banks' exposure to a fall in the exchange rate, the BRSA monitors banks open positions--i.e. their exposure to exchange rate risk--as a priority. Open positions fell to almost nothing in February, but have come back up to around $1.5 billion recently--still not a large number in relation to the size of the Turkish banking sector. BRSA Chairman Bilgin once told us he does not worry about banks on-balance sheet exposure, since this is kept quite low, but about unreported off-balance sheet exchange rate risks that are hidden. HSBC economist Ahmet Akarli told us the latter risks can be estimated by looking at the total size of the Turkish lira swap market, which he put at about $3 billion. Akarli believes this amount of risk could be accommodated by the banking sector in the event of a sharp fall in the exchange rate. In relation to the sector's total capitalization of about $34 billion this amount of exposure would not appear to be system-threatening. -------------- Corporate Open Positions a Risk -------------- ¶8. (SBU) The principal risk to the banking sector, as well as to the economy as a whole, is widely considered to reside in the corporate sector. Until recently, it was simply not economic to borrow in Turkish Lira, given astronomic lira interest rates. If they did not borrow in low-interest rate foreign currencies, corporations either received credit from suppliers, from owners or not at all. Though no one knows how exposed the corporate sector really is, many local economists believe that Turkish corporates have taken on substantial foreign exchange risk by borrowing in dollars or euros. The business daily Referans, without explicitly sourcing its information, recently put the net foreign exchange exposure of the corporate sector at $28 billion. If the lira loses 20 or 30 percent of its value, those corporates that have taken on substantial foreign exchange risk may have difficulty servicing their loans. If enough of these corporates run into financial problems, it will have a spillover effect on the banking sector, running up non-performing loans. ¶9. (SBU) Though corporate open positions probably pose the single largest risk in a bad scenario, there are mitigating factors which will limit the most severe damage to a relatively small group of upper-middle tier companies. First, small- and medium-sized enterprises simply do not have access to borrow foreign exchange. Turkish banks are still cautious to extend credit to companies other than blue chips, having been burned in the 2001 crisis and not yet having sufficient confidence in local financial statements, especially those of SME's. At the other end of the spectrum, the giant conglomerates that dominate the Turkish economy: Sabanci, Koc, Dogan, etc. are generally considered to be in good financial shape. The CEO of Akbank, for example, told us that his bank has little outstanding credit to the bank's own group (Sabanci) because its companies tend to be quite liquid. Another mitigating factor is that some of the corporates with large foreign exchange borrowings are exporters, whose risk is cushioned by the natural hedge of their foreign exchange flows from exports. Both State Planning Organization Deputy Undersecretary Birol Aydemir and former Treasury U/S Faik Oztrak separately told us that many companies are in effect "borrowing from themselves" by repatriating foreign exchange denominated assets. With large sums of Turkish money either offshore or in foreign exchange-denominated accounts in Turkey, many wealthy Turkish individuals or corporations can borrow foreign exchange from related parties. In sum, even if the $28 billion figure cited in Referans is right about the order of magnitude, the corporates at risk to a fall in exposure to foreign exchange borrowings are only likely to be a small group of companies that are both big enough to borrow in foreign exchange, weak enough financially to run into trouble, and neither cushioned by a flow of foreign exchange earnings or by the lender being a related party. -------------- Treasury Expected to Muddle Through -------------- ¶10. (SBU) Turkish Treasury's obligations continue to be heavily skewed to short-term instruments that have to be constantly rolled over. In a rush-for-the-exits scenario, Treasury will be hurt in two ways: by the effect of a depreciating lira on foreign exchange-denominated debt and by a spike in interest rates. A close look suggests that Treasury will have to make sharply higher interest payments for a time, but will manage to do so. ¶11. (SBU) Treasury has a gigantic "open position" through the large proportion of its borrowings that are denominated in foreign exchange. As of February 2005, about 39.6% of its total debt (external and domestic) is either foreign exchange-denominated or foreign-exchange-linked, down from 50.9% in September 2003. In a lira depreciation scenario, the stock of debt relative to government revenues or GDP will rise sharply. Though this will tend to dampen ministerial statements about converging towards the Mastricht criteria, it will have little effect on Treasury's ability to service its debt in the short run, because the foreign exchange debt has a much longer maturity structure and much lower interest rates than TL debt. With interest on its domestic foreign exchange-denominated borrowings currently only around 5%, after a 30% depreciation the TL amount of interest payments on this debt would still be far lower than the 17-18% interest rate it pays on an equivalent amount of lira borrowing, and could be serviced. Though a 30% increase in principal on foreign-exchange-denominated debt is significant, the FX debt--even domestic FX debt--is of far longer maturity than TL debt such that in the short run there would be negligible impact on the level of principal payments. ¶12. (SBU) As demand from foreign buyers evaporates, TL interest rates will spike. The spike in yields on government securities in secondary markets would have no immediate, direct effect on Treasury--the problem would be the 40% of Treasury debt in floating rate instruments. A spike in rates just before a quarterly redemption, for example, would be costly. On the other hand, as with FX debt, Treasury benefits from these floating rate instruments' much longer maturity, such that little principal has to be repaid in the short run, and generally lower interest rates than on discount bonds. ¶13. (SBU) Treasury is likely to be able to make its interest payments, and to deal with reduced appetite for new issuances for several reasons. First, access to foreign exchange is not at issue: the Central Bank has built up record foreign exchange reserves. Even if the Central Bank's reserves were inadequate, the floating exchange rate regime allows permanent access to foreign exchange--you just have to pay a market-clearing price. In terms of the Treasury's own reserves, Director General for domestic debt Tulay Saylan says Treasury targets a reserve level of one-half of upcoming redemptions. Treasury does not try to accumulate more than this amount because it is difficult and expensive--the Central Bank pays no interest on Treasury deposits. To deal with any shortfalls that arise if the reserves plus no issuances can't cover redemptions, Saylan has control over all state spending. In other words, if she has to, she can slow down state payments temporarily to deal with cash shortages. As for the likelihood of the fall off in demand for issuances to be so severe, that she simply cannot fund redemptions, note that even in corrections such as the April-May 2004 sell-off Treasury has been able to issue new paper. Turkish banks always need a supply of this paper: though they are developing their lending business, it still cannot cover Turkish banks' deposit base, such that banks need government securities in which to invest their deposits. ¶14. (SBU) In 2005, Treasury has taken advantage of favorable market conditions to issue more long-dated paper, thereby reducing its need to constantly roll over such a large share of its debt. Tulay Saylan told us that the average maturity of new domestic issuances for the first quarter of 2005 is 2 years, up from only 14 months in the same period last year. The average maturity of all outstanding domestic debt as of February was 21 months. For the first time in memory, Treasury issued a domestic, TL-denominated 5-year bond in February, coming on top of a path-breaking 3-year issue in October, 2004 (since repeated). One factor that has helped Treasury, by deepening the market for long-dated TL Treasury paper is foreign banks' issuance of TL-denominated debt. A portion of the proceeds of the foreign banks' issuances is reportedly invested in long-dated Turkish Treasuries--with over $4 billion of these issuances in recent months this is a significant new source of demand. -------------- Longer-term effects -------------- ¶15. (SBU) The ability of Treasury and the banking sector to muddle through in the short run does not mean there is no risk of disruption to the economy and the GOT's efforts to strengthen its financial position. If the lira fell 20 or 30 percent and stayed at a new lower-level equilibrium for a sustained period there is likely to be substantial disruption in the "real economy." A recession would likely result from a series of sudden shocks: suddenly-expensive imports would stress importing companies, bring back inflationary pressures, and hit the man on the street with a double-whammy of overnight reduced purchasing power and job losses. Stress on importing companies and low- and middle-income Turks could engender a new wave of difficult political problems. Instead of a short-run, financial crisis requiring a financing package, in this scenario there would be a slower-developing but still potentially destabilizing social-political crisis. The resurgence of inflation pressures would postpone the final success of the Central Bank's disinflation campaign, but the Bank could deal with price pressures by tightening up monetary policy. One Istanbul analyst said that if such a scenario occurred, it would be the first real test of the Central Bank's mettle. It would also test the Government's shaky commitment to the independence of the Central Bank. ¶16. (SBU) The effect on fiscal policy could be more dramatic. The GOT would have to further postpone its dreams of consituency-placating expenditure. Tax revenues are highly pro-cyclical, i.e. they tend to overperform in a strong economy and underperform in a slowdown. This is particularly true in Turkey's case, given that 70% of its revenues derive from indirect taxes like VAT or other consumption taxes. The GOT will be extremely hard-pressed to maintain fiscal austerity in shock-induced recession. -------------- Comment: Anchors Help -------------- ¶17. (SBU) The above analysis applies regardless of the how the GOT is doing with the IMF and the EU. Even if things seem stalled with both, a crisis can probably be avoided in the short run. To the extent the GOT is seen to be moving forward with one or the other anchor, it will dampen the severity of the disruption. Moreover, if the GOT is not moving forward with either the IMF or EU when the bad scenario develops, the authorities' past behavior suggests they will suddenly be galvanized to market-pleasing action, as they were in March 2003. ¶18. (SBU) But we should not let this analysis lead us to match the GOT's complacency with our own. The GOT's half-hearted commitment to economic reform needs to be constantly encouraged. Though Turkey's vulnerability to a purely financial crisis has declined, good policies are vital to cushion the severity of the fall of the exchange rate in a shock scenario. EDELMAN