Brad Setser

Brad Setser: Follow the Money

Print Print Email Email Share Share Cite Cite
Style: MLA APA Chicago Close

loading...

Levered long/ long funds …

by Brad Setser
June 29, 2006

An anonymous partner at a mid-sized London hedge fund expressed something I was trying to say a couple of weeks ago fair better than I could have.   From the Financial Times:

"A lot of managers have gone from being long-short funds [funds that make bullish and bearish bets on stocks] to being long-long funds with leverage, which removes the whole point of hedge funds offering protection in the event of a downturn," said a partner at a mid-size hedge fund in London. 

I expressed my argument somewhat less elegantly – saying that that some hedge funds were not really hedged.   And many of my readers pointed out  — quite correctly – that no one that is fully hedged makes money.  

But what I was getting at was that it was quite costly to hedge say a long position in an emerging market equity market with an offsetting short position in the same equity market.    Funds that hedged in the same market didn't do as well as funds that did not hedge.    Until May, a rising tide was lifting all boats.  And punishing shorts. 

So there was a temptation to become a long/ long fund.

Or to find proxy hedges that didn’t cost you an arm and a leg. 

The one that I am most familiar with comes not from emerging market equity markets, but from the market for the local currency debt of emerging economies. 

It went like this: 

Buy the local currency debt of an emerging economy with a high coupon.

Don’t hedge the currency risk.  That cost you. 

Instead buy insurance against the risk that the country would default on its foreign currency denominated debt.   That meant holding a credit default swap – i.e. buying credit insurance.  

This trade isn’t a secret.  Joanna Chung, quoting Mohammed Grimeh in the FT

Mohammed Grimeh, global head of emerging market trading at Lehman Brothers, said: “Over the last few months, as credit spreads tightened, hedge funds have been moving to illiquid local instruments that offer a higher risk and reward but they are also buying CDS protection against overall country risk.”

The IMF also discussed it in its most recent financial market update (see footnote 4 on page 9).

It isn’t hard to see the logic of the trade.  Turkey was paying a 12% coupon – maybe a bit more – on its local currency debt at the beginning of the year.  Insurance against a default on Turkey’s dollar bonds cost maybe 200 basis points.   

You were hedged … sort of.    And the whole trade had a nice carry.    A real nice carry.   After all, you were holding a rather overvalued currency and needed some compensation for the risk. 

It was a long/ short trade, not a long/ long with leverage trade.  But the long was on local currency debt and the short was on external foreign currency debt.

The hedge even has worked in Turkey.   Sort of.   A contract insuring against the risk of default on Turkey’s foreign currency debt is worth a lot more now than it was in January.   Unfortunately, both the Turkish lira and Turkish lira bonds are worth a lot less now than they were then too. 

Still, it has struck me that there was something a bit strange with the trade. 

It worked in a backward looking sense.   In the past, indebted countries often had lots of local currency and foreign currency debt.  So when a country’s currency has declined and the value of its local currency bonds fell, the value of the country’s foreign currency bonds also fell.  Credit spreads rose.  So the value of a contract providing insurance against default rose. 

It actually worked in another sense.  In times of huge distress, the local currency debt of some countries is arguably less risky than the foreign currency debt.  Countries like Argentina defaulted on their foreign currency debt but honored their local currency debt.    So in a really bad scenario, it arguably made sense to be long the local currency stuff and short the foreign currency bonds.  Never mind that Russia did the opposite of Argentina, defaulting on its local currency bonds and honoring its Russian era Eurobonds.

Incidentally, the other trade that works in backward looking stress tests is “long the external bonds of Argentine corporates, short the external bonds government of Argentina.”  Both firms and the government defaulted.  But you did better in the restructuring holding the bonds of the Telcos than the government …  

So why do I say that there was something strange about the hedge?  Simple:

The hedge implicitly assumes that correlations that held in the past would hold in the future, even though conditions change.    

And in this case, two conditions were changing:

First the scale of foreigner’s local currency exposure was growing, big time.  That changes the country’s incentives. 

And second the amount of sovereign foreign currency debt was falling, big time.   Still isChris Mewbourne of Pimco thinks repayments will top new issuance this year.  So the stock of external debt is falling absolutely, not just relative to sovereign reserves.  That is why insurance against the risk of default is cheap.   

At the limit, a country might not even have any external debt left to default on.   Or just a trivial amount. 

But a bank could still sell insurance against external default.  Why not?  The risk is low.  And nothing prevents the folks from selling more insurance than there are bonds.  Delphi showed us that.  Selling insurance (off the balance sheet) is another way to get an easy yield pickup.

And folks holding local currency debt could still buy insurance against default on non-existent foreign currency debt as a hedge. 

The backward-looking correlations work.  And why worry too much about the possibility that the conditions that gave rise to those correlations have changed?

Sorry about an obscure post.   I am sure that my tendency to delve into this kind of thing (and ferret out obscure reserve data) is one reason why this blog isn’t always an easy read. 

But the operation of the sovereign debt market is one of my long-standing interests.

As are market dynamics under stress. 

I suspect that a lot of folks scrambled to find hedges for previously unhedged positions over the past two months — and to shore up their proxy hedges.    But I am reading market tea leaves, not speaking from direct knowledge.

And I do still wonder about the robustness of some of the hedging strategies used by players in the credit markets of advanced economies in truly adverse conditions …

26 Comments

  • Posted by Guest

    The liquidity question: Liquidity can also play havoc with measures of asset correlation. Two assets can appear to be uncorrelated, making them ideal components of a diversified portfolio. But investors who spot that lack of correlation will start to buy the assets simultaneously.

    The correlations will thus increase as assets flow into the sector. And if both assets are illiquid, then the correlations will get even higher when investors start to pull out again. The rush for the exits will be all the greater if investors have used borrowed money; they will simply have to sell to cover their debts.

    When the subject of painful deleveraging is raised, the mind inevitably turns to hedge funds. . Some funds do not allow investors to redeem their holdings for long periods (a year, or even five). But most investors access the sector via funds of funds, which offer much more frequent liquidity.

    If funds of funds face redemptions, they will be unable to meet them by selling funds with long lock-up periods. They will thus be forced to sell their most liquid investments. The managers of those underlying investments will suffer, particularly if the rest of the market is aware of their problem; other funds will trade against their positions. There accordingly could be a “race to illiquidity” with both funds and funds of funds trying to lock up investors for as long as possible.

  • Posted by Guest

    The process will reduce the attractiveness of hedge funds as a sector: On the same theme, the SEC, the US regulator, warned this week about “side letters” issued by hedge funds, some of which give preferred investors the right to sell first. It is bad enough paying high fees to get into a hedge fund; it would be pretty galling to find you were a second-class investor as well.

  • Posted by Guest

    So, now is a hedge funds message board?

  • Posted by wcw

    Isn’t saying, “[t]he hedge implicitly assumes that correlations that held in the past would hold in the future, even though conditions change” the way every modern HF-alike tragedy begins?

    There’s nothing wrong with using indirect or proxy hedges, until there is. It’s like trendfollowing, which works, until it doesn’t. Both can be pretty painful for the folks engaging in them; betting on reversals or regime changes can be extremely trying, until it finally pays off.

    What can I say, timing is tough. I can’t do it; you should see my ’99 returns. Ouch.

  • Posted by STS

    Brad:

    “Sorry about an obscure post. I am sure that my tendency to delve into this kind of thing (and ferret out obscure reserve data) is one reason why this blog isn’t always an easy read.”

    Nothing to apologize for! If I wanted an easy read, I wouldn’t follow econ blogs at all. Insight into the thinking of hedge fund managers is very useful, and germane to your main topic as well.

  • Posted by bsetser

    “There’s nothing wrong with using indirect or proxy hedges, until there is. It’s like trendfollowing, which works, until it doesn’t. Both can be pretty painful for the folks engaging in them; betting on reversals or regime changes can be extremely trying, until it finally pays off.”

    wcw — well said. I certainly learned how trying betting on a reversal or regime change that does happen can be last year. In a small way financially. And in a big way reputationally.

  • Posted by psh

    ‘trade that works… long… corporates, short… government bonds. Both… defaulted. but you did better in the restructuring holding [Telco] bonds’
    That may be something of an eternal verity. CDS value is not so sensitive to the recovery rate (because implied default probabilities vary directly with the recovery rate, offsetting payoffs which vary inversely with the recovery rate) So there may be some incremental value in keeping the debtors you can push around but getting protection for the government debt. Do it the other way, and you’re not duly rewarded for all your hard work hanging deadbeats out the window by their feet.

  • Posted by dryfly

    I second STS – if I wanted shallow & easy to digest information I’d get all my news from Fox News & MNBC etc.

  • Posted by jm

    … hedge funds have been moving to illiquid local instruments that offer a higher risk and reward but they are also buying CDS protection against overall country risk.

    And in how many cases is it other hedge funds selling them the CDS protection? As the quote says, “Selling insurance (off the balance sheet) is another way to get an easy yield pickup.” While that was referring to banks as the sellers, it’s not necessarily limited to them. As long as there’s no major upset in the markets, it’s “an easy yield pickup”. But it’s at most just a few percentage points of the worst-case liability being taken on. As the “names” at Lloyd’s of London discovered some years ago, insurance can change from lucrative business to bankrupting disaster if a “fat tail” event brings in a flood of claims.

  • Posted by Guest

    “…According to the latest housing survey from the Nationwide building society, the average house now costs nearly £166,000 with mortgage repayments absorbing 42% of average take home pay… The [Council of Mortgage Lenders] CML believes home ownership has not reached any natural limit and expects both home ownership and mortgage borrowing to continue expanding. Mortgage debt has also been boosted by the enthusiasm of home owners for borrowing against the increased value of their homes, known as mortgage equity withdrawal. More than £200bn has been borrowed this way since the start of the decade…” http://news.bbc.co.uk/2/hi/business/5128220.stm

  • Posted by JCK

    “Selling insurance (off the balance sheet) is another way to get an easy yield pickup.”
    usually,there is no free lunch…but in this case there is if the cds is written by a local bank , if there is no default they pick up the premium,if there is a default your insurance is worthless so hedging turkish lira debt with cds on turkish $ debt is a particularly stupid idea…

  • Posted by Guest

    “And in how many cases is it other hedge funds selling them the CDS protection?” A hedge fund that buys CDS protection from another hedge fund deserves to fail, and probably will.

  • Posted by Guest

    Not sure how or why economists got into the prediction and market timing business when telling it like it is seems to be enough of a challenge. Once those cards are revealed, don’t market participants immediately organize and act on their own strategies (bets?) to profit from, and alter anticipated outcomes?

    “…We can debate for hours the nuance of the Fed statement. That is a job for others. My job is to get on board the right trades at (roughly) the right time…” http://www.fxa.com/plant.asp

    “Corporate America is about to witness a sharp rise in bankruptcies caused by the recent boom in debt-funded acquisitions and hedge funds’ growing appetite for take-overs, according to Wilbur Ross, the veteran investor in distressed businesses. Mr Ross, who has accumulated an estimated $1bn fortune by investing in troubled industries such as textiles and coal, said the highly-leveraged buy-outs of the past few years would take their toll on several US companies. “I think we are going to see a big increase in bankruptcies in 2007-08,” Mr Ross, chairman and chief executive of the private equity group that bears his name, told the Financial Times. “Every time you compound leverage, you cut the margins for error.”…” Wilbur Ross predicts rising bankruptcies, June 29, 2006, ft.com

    Off topic, but if you haven’t come across it, the following in the current issue of Grant’s may be of interest:

    THE NEW OLD JAPAN – “The high-speed Chinese economy is bound to decelerate. Sooner or later, the restrictive policies put in place by the People’s Bank of China will take hold, Sinologists and economists broadly agree. Something will do the trick, they insist: If not the boost to the basic lending rate on April 27, then the tweak to reserve requirements on June 16 or the program of intermittent sales of central bank securities… For ourselves, we don’t believe it. We don’t doubt that the Chinese economy will eventually throttle down. But monetary tinkering will avail the authorities nothing until the renminbi-dollar exchange rate is untethered. The essay now unfolding is, however, more than a brief for a higher renminbi and a lower dollar. It is also a speculation on the consequences of the Chinese style of finance to the world’s economy and markets…”

  • Posted by bsetser

    Tis true, sellers of protection are not limited to banks — tis also an easy yield pickup for hedge funds, and even some real money funds. PIMCO said it was selling protection for countries with tons of reserves a while back.

    As for why economist go into the prediction business, well — some of us don’t work for academia, and most folks in the market aren’t interested in knowing that the trade deficit is growing, or even that the income balance will tip into deficit (barring a surge in dark matter). They want some indication of why/ when something might come along to change this “stable” disequilibrium.

    And in my case, I have always focused on criseconomics and how extreme market moves impact on economies. It started with Asia. And I then spent a lot of time working on Argentina. And now I am interested in Iceland and Turkey as well as the difficult to explain/ understand case of China (extreme, but in another way).

    How does that all relate back to selling CDS to hedge v local currency risk? Simple: the big change over the past few years has been the growth in international local currency exposure, and I want to understand how this might change market and economic dynamics in times of stress.

  • Posted by MrBill

    Hedge Fund always was a bit of a misnomer. They have long gone from being hedges against bearish moves in equity markets to being simply another form of mutual fund using any number of strategies, long, short, liquid, illiquid, etc.

    For private equity funds, the secret is to keep the prospectus quite general, so you do not tie your hands if market conditions change. Yes, you have to have a track record so investors trust you with a blank cheque, but say emerging markets suck at the moment, you need to be able to move into the next big thing without going back to your original investors and changing your mandate. Another approach is to have an umbrella fund and then set-up sub-funds with different mandates. That works as well.

    Local government debt and foreign denominated sovereign debt are substitutes for one another. It really depends on liquidity and investor appetite, and the government may wish to establish a benchmark even if they do not need the money just so that corporates can also issue debt more cheaply because investors have that benchmark.

    Historical spreads as you point out depend on the amount of outstanding debt in absolute terms or as a percentage of GDP. So any funds buying protection based on historical data may have missed that many emerging market issuers have paid down their external debt. Russia for example. Also, domestic inflation may be high, but as a foreign investor you may not care, if the carry trade is still there. Again Russia for example, where the ruble is appreciating.

    Even Turkey has been hit less hard in this recent sell-off than in previous lira crises as they have learned some lessons from the past. So using only historical correlations is a dangerous game.

    Of course, in this last sell-off S. Africa, Turkey, Hungary and others with current account deficits were hit harder than Russia, for example, but as they are still grouped together in the same asset class, Russia also got sold off despite its better fundamentals. If you have to make margin calls, you have to close profitable positions as well. But fortunately, Russia was the fastest to recover since then as once the panic is past investors head for the better fundamentals and bargain hunters look for cheap equities and bonds.

    I think it is ironical that in the great banking consolidation, the trend was to shut-down small dealing rooms and integrate them onto ever larger trading floors, but as performance diminished, these banks farmed out a lot of prop trading to hedge funds, often staffed by ex-employees. True pay for performance and lower fixed costs. But in the world of hedge funds, sometimes small is beautiful.

  • Posted by MrBill

    Debt in local currency can also be hedged using an NDF instead of a CDS as if the credit quality goes down chances are the foreign exchange rate will weaken as well. Plus it is easier to trade in and out of NDFs vs. CDSs quickly. You can even take protection using futures on the CME likely. Cheers.

  • Posted by MrBill

    “”Selling insurance (off the balance sheet) is another way to get an easy yield pickup.”
    usually,there is no free lunch…but in this case there is if the cds is written by a local bank , if there is no default they pick up the premium,if there is a default your insurance is worthless so hedging turkish lira debt with cds on turkish $ debt is a particularly stupid idea…””

    Double idemnity risk or highly correlated risk between underwriter and the risk you are trying to hedge is why most banks would not buy Turkish default risk from a bank in Turkey. 99% correlation.

    However, ironically in the Russian Crisis in 1998, there was a two tier market where you could buy NDFs from a Russian counterpart, or spend more money and buy NDF protection from a US bank. Needless to say, when Russia defaulted, the US banks declared force majeure and the offshore insurance was just as worthless as the local insurance that was declared null & void by the Russian courts.

    So even if you buy a CDS or a NDF from Citibank or JP Morgan and you’re sure they would not want the reputational risk from a default, don’t kid yourself. When millions and hundreds of millions are on the line their lawyers will be looking at every angle to reduce their payout and preserve shareholder value. Call it Legal Risk and add it to your Default Risk as well as Country Risk.

  • Posted by Joseph Wang

    Part of the interest in “weird stuff” is because returns on the “non-weird stuff” have been so lousy the last few years.

    One of the interesting things about derivatives is how it requires some pretty strong lawyering. When you end up losing hundreds of millions of dollars (and it happens) you are going to look for any possible loophole or legal argument that you can find to get out of those losses. so this requires a lot of upfront lawyering to insure that the contracts are bulletproof.

  • Posted by Guest

    re: “indication of why/ when something might come along to change this “stable” disequilibrium”

    I see this more as speculation and scenario building than prediction. The determination of possible outcomes given what is known midst a whole bunch of complexities and unknowns. Ultimately it’s up to those taking the positions to plug in the pertinent details and place their bets. The results of those collective actions alter the baseline and scenarios, taking the game to another level.

    re: “using only historical correlations is a dangerous game.”

    Agree – it really is different out there.

  • Posted by Guest

    re: “One of the interesting things about derivatives is how it requires some pretty strong lawyering”

    Along with lots of innovative accounting?

  • Posted by Guest

    And along with the lawyering and accounting, one heck of alot of IT.

  • Posted by bsetser

    Didn’t know Citi/ JPM used Russia’s capital controls to get off the hook. The Russian banks would have defaulted. capital controls or not. Interesting. you learn something everyday. Interestingly, I think the Spanish banks honored the CDS sold by their local Argentine subs in 00/01. Correct me if i am wrong, though.

    Yes, you can hedge currency risk in the NDF market. My sense though is that if you hedged your say turkish lira risk using NDFs, you had to pay a decent price. So basically you unloaded the currency risk (and currency upside) and were mostly betting on a fall in long-term lira interest rates. It was a directional bet on the lira denominated bond market, without the lira risk, just the local market interest rate risk … at least that was my sense. I never worked through the math.

    Nice discussion, by the way.

  • Posted by Guest

    according to grant interest in “weird stuff” has increased because it’s harder to earn a commission since TRACE was implemented in 2002…

    http://www.bloomberg.com/apps/news?pid=10000103&sid=abuggIP.aitE – “The system now provides prices and the amount of bonds exchanged in each trade on 29,000 securities within 15 minutes of a deal, according to NASD. With the data available immediately, there’s little need for guidance from analysts or salespeople.”

  • Posted by Joe S. Pack

    Trying to figure all this out hurts my head – which is probably why I’m still a working stiff. CDSs and CDOs are just another LTCM acronym to me…a big old snowball rolling down a hill that replicates itself whenever the responsible party(s) is unable to pay it (them) to stop.

    International bond funds typically advertise themselves as hedged and unhedged. I have an IB fund that has a bunch of stars and performed admirably between 1994 and 2002. During this period they were fully hedged, so the returns were consistent and quite good and the volatility was low. In 2003 they went unhedged – they must have been reading Brad’s blog and/or were envious of the large dollars that the unhedged IB funds earned over this period.

    So then we started getting ¼ point increases on a regular basis. Death by a thousand cuts as the carry trade obscured the Wile E Coyote floor of the twin deficits and the dollar kept creeping up. I have to be satisfied with getting the benefit of the relative security of 1st world sovereigns in countries with adults running fiscal and monetary policy (if not always in tandem). I look at the more volatile, flat-ass performance of my unhedged fund for the last couple of years and go – either my Keynesian macro education was absolutely worthless or the dollar is going to head for the value of the basement where it belongs. Then I’m reminded of Maynard’s long range projection on the probability of ending up in the nitrogen cycle. All this by way of saying, there ain’t many “swaptions” in my IB 6-month financial statement, and for this at least, I am grateful.

  • Posted by Guest

    re: “…the data available immediately, there’s little need for guidance from analysts or salespeople..”

    But an increasing need for people who can structure and interpret the data.

  • Posted by MrBill

    Brad wrote””Yes, you can hedge currency risk in the NDF market. My sense though is that if you hedged your say turkish lira risk using NDFs, you had to pay a decent price. So basically you unloaded the currency risk (and currency upside) and were mostly betting on a fall in long-term lira interest rates. It was a directional bet on the lira denominated bond market, without the lira risk, just the local market interest rate risk … at least that was my sense. I never worked through the math.””

    Sorry, I did not mean to imply that you could hedge out your currency risk using an NDF AND still benefit from the interest rate differential between the currencies. If you can, there is an arbitrage or market inefficiency in the pricing.

    However, I assume you go into Turkey (or Russia) on T + 2 thinking it is a good yield pick-up story and you’re comfortable with the country’s fundamentals, or you would not likely put the trade on in the first place.

    But let’s say at T + 180 days you’re locked into a one year trade and you’re starting to get nervous. Then from a market timing perspective, you can buy your NDF to hedge the currency exposure or a CDS for the credit protection, AND then you’re only paying for 6-mos. protection versus one year.

    I also assume you’re buying protection when vols are low and the sun is still shining and not looking for the proverbial umbrella (and a taxi) after it starts raining? And on that front, what a difference a few weeks can make (he says, as Russian equity bounces +28% from its sell-off in June)! ; – )