Wednesday, December 15, 2010
Although I claimed it would be unwise to do so at the end of my latest Roubini Global Economics contribution (updated version of this week's Hurriyet column, although the updating only changes one word, "credible" to "wise" in the last sentence, as the weekend presentation has made it credible all right), it looks likely that the CBT will be cutting the policy rate tomorrow: Vice Governor Erdem Basci, noted as the most likely successor to Governor Yilmaz in a U.S. Istanbul Consulate WikiLeaked cable, in an emergency presentation at the Turkish Economic Association, clearly spelled out that a policy rate cut would be on the agenda at tomorrow's MPC meeting.
The overheating argument I spelled out in the Hurriyet/Roubini columns is only a part of the picture. To see what is going on, it is important to understand the Central Bank's rationale before getting into a discussion on whether this will work and the potential Turkish asset price consequences.
What is the Bank's rationale:
In one sentence, to deter capital flows and address the current account deficit & loan growth. Note that a casual look at the latest Balance of Payments, or BOP, statistics tells you that these two issues are intertwined: The most significant contribution to the October capital account, i.e. financing of the current account deficit, was from short-term bank borrowing. Turkish banks borrowed net $2.1 billion in short-term loans from abroad in October, roughly the same amount as in September and a record-number. Those figures look really nice when graphed alongside loan growth:
So you know where the banks' loan growth is coming from...
Note that cutting the policy rate is only part of the picture: In his presentation, Mr. Basci is hinting at a two-tier strategy: The idea is that short-term lira rates could be decreased to deter lira appreciation, while at the same time hiking reserve requirements for banks in order to rein in credit creation.
And what's the rationale for their action? To (liberally) translate slide 5 of the presentation: "Increasing the policy rate, while it would reduce the current account deficit from the credit channel, it would have the opposite effect from the currency channel. It is not possible to halt the widening current account deficit via short-term rates, the basic policy tool of the Central Bank, alone. Therefore, tightening non-interest tools to curb credit growth, while at the same time decreasing the policy rate in a controlled fashion to limit currency appreciation, has emerged as the ideal policy response to the widening current account deficit." (footnote to my perennial spammer: you win! Once again, I have no analysis, just quotes)
Mr. Basci's presentation contains more details: As their goals, he notes that they would like to lengthen the maturity of both deposits and capital flows as well as a steeper curve in deposits and swaps. In addition, he is explicitly saying they would want to make short-term swap rates more volatile than long-term swap rates, to discourage carry-traders, just as they had done when they cut the overnight borrowing rate.
Will it work?
It is possible, but I personally don't buy it. First, it is unlikely that the hot money flows will be too much affected by a 50 or 100 basis points cut in the policy rate, even though the rising U.S. Treasury rates are working in the Bank's favor of late (sorry for the double axis, but it makes my point much clearer):
On the contrary, it is equally likely that that the falling rates will cause a rally in Treasuries and equities, attracting even more hot money. And the other details of the Bank's great current account reduction scheme are not without problems, either. For example, with the average maturity of deposits only a couple of months, it is also equally likely that the rate cut would reduce the marginal propensity to save rather than channel deposits to longer maturities, as the Bank is hoping, worsen the country's private savings problem and widen the current account deficit further as a result.
But let's give the Bank the benefit of the doubt and assume that it managed to discourage short-term capital flows and weakened the lira, with the current account deficit getting under control. It is not given that credit growth will automatically dampen, or that the Bank's measures towards that goal will work.
My skepticism arises mainly form my firm belief that to reduce the quantity consumed of any good, its price will ultimately have to go up. In our case, to curb credit growth, its price, the interest rate, will have to go up. Reducing the policy rate, even with the accompanied reserve requirement hikes, could easily feed its way to consumer confidence, more consumption and eventually to the current account deficit.
The Central Bank's proposed actions also create an interesting dilemma of tightening the quantity of money and reducing its price at the same time. If the Bank cuts downs the one-week repo rate but does not provide enough liquidity at the repo auctions, then market rates will go up. And if it will, then the extra liquidity will find its way to credit. Of course, that's the idea behind simultaneously hiking the reserve requirements, but still...
But speaking of money and liquidity brings me to a related dilemma: I am not going to get into whether inflation is mainly a monetary phenomenon or not, but the rapid growth in the Turkish monetary base does not need to be debated. This arises from two factors: 1. The growing role of net foreign assets, i.e. capital flows; this is the Central Bank balance sheet way of seeing the capital flows working their way into domestic credit, as I discussed above. 2. The Central Bank's reluctance to lower its liquidity support and fully sterilize FX interventions, as the picture below shows:
Why is the Central Bank not doing so? Simple: If they did, they would create an upward pressure on interest rates. But then this would be inconsistent with the policy strategy outlined in Erdem Basci's presentation. But then again you are essentially easing financial conditions with this excessive liquidity support. See the dilemma?
Enough of dilemmas... Just for the sake of argument, let's go one step further: Let's assume that the Bank's policy strategy manages to curb credit growth as well so that there isn't a lot of pressure in inflation from that front. But what about the the effect on inflation because of the now-depreciated lira and sticky expectations? As economist Atilla Yesilada notes at his Turkish economics analysis website Bilgeekonomist (i.e. the wise economist, highly recommended if you speak Turkish), there could easily be ripple effects such as negative deposit rates, booming consumption, double-digit bond rates and higher interest expenditures.
Now, you probably see why I am worried about the Bank's new course: It depends on too many "if"s... In essence, the Bank is behaving more like an ardinal (adrenaline a la Turca)-junkie trader rather than a conservative Central Bank. But then again Yapi Kredi Chief Economist Cevdet Akcay, one of the few supporters of the Bank's new strategy, argues in his latest weekly that these are extraordinary times, and extraordinary times call for extraordinary measures. He may be right, but again: Such risk-taking should not be a Central Bank's cup of tea, especially as markets have made fool of central banks all the way from the United States to China, as Mr. Yesilada notes in the same piece mentioned above.
An important side point
While it does further explain the Central Bank's rationale as well, this side point is also a discussion of the overheating argument of my most recent Hurriyet/Roubini columns and is a bit technical, at least conceptually. But feel free to skip it if it is not your cup of tea.
Note that an important point of the Central Bank's argument, which I have ignored until now, is that the economy is not overheated. In fact, they have the following nice chart, from the summary of the Financial Stability Report (also in Mr. Basci's presentation), to illustrate their point:
Unfortunately, there is no English version, so let me summarize what it means: The x-axis is financial stability, with quadrants of slowing and rising credit growth. The y-axis is price stability, where the quadrants are slowing and rising inflation. For example, the Southeast quadrant illustrates an economy where credit is growing, but inflation is slowing down. The Bank thinks we are in the originating bold oval, i.e. it sees credit growth faster than desired and inflation about where the Bank wants it to be. Originally, the Bank was seeing the economy move towards the Base Scenario, towards the Northeast, but now it sees the economy moving towards Scenario I. Whereas the Basis Scenario would require tightening both the policy rate and other tools, Scenario I calls for tightening other tools, while at the same time, cutting the policy rate.
As I explained in my latest column, I am already seeing signs of overheating in the economy, and so would argue we are a bit to the North of where the Bank says we are. And I would also argue we are heading to the northeast, i.e. more inflation and credit growth, i.e. the Bank's Basis Scenario.
The key problem here is with the difficulty of measuring the degree of overheating in the economy. In fact, Cevdet Akcay and his team just published an excellent, albeit slightly technical, note explaining exactly the same point. In an email exchange, Eren Ocakverdi, one of the authors of the report, noted in an email exchange that using multivariate filters, like the ones your friendly neighborhood economist loves to use, to calculate Turkish potential output is problematic. Therefore, he argues that, without a reliable supply-side indicator, output gap estimates based on potential output, like my projections!, are not reliable. He further emphasizes that their note was written to illustrate this fact rather than claim that their preferred filter is superior to the commonly used H-P. I see his point, but just the fact that I have tried three slightly different different methodologies so far and always come up with similar results (almost-closed output gap) should mean something -and not only that I am econometrically-challenged, more than that:)...
What would I have done?
According to your favorite Turkey economist, the ideal response would have been minor monetary tightening, major fiscal tightening, some macroprudential measures like higher provisional requirements for new loans and the sterilized FX interventions I have discussed above. As I have argued before, I would expect these measures not only to curb credit growth, but also bring down nominal interest rates and slow down hot money this way. If that doesn't work, a Tobin tax (or the like) could be the next step.
Note that it is very easy to fire away from the sidelines. Just as Turkey has 60-odd million football (that's soccer for all the Yanks) managers, who love to play the role their favorite team's manager every week, the country also supports quite a few economists (or economist wannabes), whose favorite pastime is second-life Central Banking:)
First of all, anyone with a few economics courses under her belt would have come with my prescription, but theory and practice are very different animals: For one thing, the macroprudential measures I am mentioning are at the reign of the banking regulator, i.e. BRSA. As for fiscal restraint right before elections: Forget about it... So maybe, as Murat Ucer of Turkey Data Monitor was noting a couple of days ago, since the Bank knows there is no way it is getting support from the government (fiscal) and BRSA (remember the BRSA's row with the Bank right after the latter's capital punishment on the banks), it is merely implementing a second-best policy response. And crossing fingers, praying and hoping for the best...
Also note that the Bank is acting very Machiavellian: With Turkey's risk premium very low at the moment and inflation to fall to as low as 5 percent in the next three months, mainly because of base effects, if the Bank is to cut the policy rate, this may be the best time. In fact, if U.S. interest rates continue to remain elevated, the Bank may see its actions bear fruit on the capital flows front sooner than expected. My only worry about the timing is that it is the end of the year: With traders down in Switzerland for Christmas skiing, markets will be thin, so we may see sharp moves in either direction...
Implications for Turkish assets:
Again, first the executive summary for the time-constrained: Weaker lira and a steeper yield curve, exactly what the Central Bank is shooting at. And that's what we got in the first half of the week:
With stocks, it is a bit more complicated: First, banks are the dynamos of the Turkish stock market, and they are to be faced with two opposing effects: The rate cut (+) and reserve requirement hikes (-). So at the end of the day, the size of these measures will matter. Moreover, there are many unknowns to the equation, such as the response of the yield curve, the response of credit demand and volatility in the lira going forward, and last but not the least the path of global interest rates (and what it means for the global economy- Martin Wolf has an insightful piece on this).
A word on communication:
Mr. Basci's last slide is titled stability and communication. I am not sure what he explained there, as I was not at the presentation, but the slide refers to the Bank's periodic communication tools, i.e. the Inflation and Financial Stability reports, as well as a couple of recent speeches by Governor Yilmaz.
While I totally agree with Mr. Basci on the relationship between stability and good communication, I don't think those reports or a last-minute weekend gathering would exactly classify for good communication. It seems that the Bank decided on the weekend presentation when its rate-cut signals did not go down as much as it had hoped to and wanted to relay the message straight this time around. But the message did not get through the first time exactly because it was going through the wrong medium! MPC meetings/ summaries and scheduled (i.e. pre-announced) speeches should be the medium for that. In that respect, I am looking forward to the scheduled speech by Governor Yilmaz in Konya this Friday.
There is quite a bit of gossip in the media and markets that all this is as much about the race for the Central Bank governorship as actual policy direction. The current governor Durmus Yilmaz's term ends in April, and he has indicated he will retire. One of the strongest candidates, if not the strongest, for the post, is current Vice Governor Erdem Basci, the hero of the weekend presentation. It is claimed that, given that the Governor will be assigned by the government, and that PM Erdogan's inclination towards lower interest rates is well-known, Basci positioned himself well for the position with his weekend presentation. In other words, the presentation was as much about the man as the ideas.
Hurriyet Daily News, as well as many Turkish newspapers, published a piece on Monday touching the same angle. I find it hard to believe this theory, but then again, I have never been good at office politics, internal strife and the like. But one thing is for sure: If there is some truth to it and word gets out that the MPC is divided on such an important policy change, it would hurt the Bank's credibility real bad...
Then, the hero of the story might find out that he really needs to be saved (listen to the lyrics) after all...