from Follow the Money

Relitigating 1998 at the end of 2008

December 31, 2008

Blog Post
Blog posts represent the views of CFR fellows and staff and not those of CFR, which takes no institutional positions.

More on:

Financial Markets

Tyler Cowen argues that the “Committee to Save the World” made a mistake in 1998 by, well, saving the financial world. They thus missed an opportunity to teach the banks a lesson in sound risk-management.

He specifically argues that the Fed (actually the New York Fed) shouldn’t have called the big banks together to recapitalize LTCM. The recapitalization didn’t require any Treasury funds or draw on the Fed as a lender of last resort, so calling it a bailout obscured the meaning of the term bailout – the fed catalyzed a private bailout of LTCM but it didn’t do a true government bailout. You might even say that the Fed catalyzed a bail-in of LTCM’s creditors. But by acting, Cowen argues that the Fed set a precedent that creditors of big financial institutions don’t take losses, and thus encouraged bad bets.

I am not totally sure. The big banks called to the New York Fed were the creditors of LTCM and they were in some sense “bailed-in.” To avoid taking losses on the credit that they had extended to LTCM, they had to pony up and recapitalize LTCM.*

It just so happened that the market recovered and it was possible for LTCM to exit many of its positions without taking large losses, or in some cases any losses. The banks that took control of LTCM when LTCM was on the ropes were able to unwind LTCM’s portfolio in a way that didn’t result in additional losses. But the result Cowen desired -- large losses for the banks and broker-dealers who provided credit to LTCM – was quite possible if LTCM’s assets weren’t sufficient to cover all its liabilities. No creditor of LTCM was able to get rid of its exposure as a result of the Fed’s actions.

Lehman’s creditors didn’t get a chance to do a similar deal. There were too many of them -- and there was too little time. I suspect, though, that Lehman’s creditors and counter-parties would be far better off if they had all agreed to pony up say 10% of the money they had lent to Lehman and in the process had provided Lehman with enough equity to allow it to be unwound in a more orderly way.** They still would have taken losses, but those losses might well have been smaller – even counting the new money they put in – than the losses that Lehman’s creditors will incur as a result of Lehman’s bankruptcy filing.

Moreover, it seems a bit strange to look at LTCM in isolation.

LTCM, remember, came just after Russia defaulted.

And Russia was at the time considered the quintessential moral hazard play.

A host of financial institutions thought it was too nuclear to fail, and thus concluded that that they could safely pocket the high coupon on Russia’s GKOs (short-term Ruble denominated Russian securities) …

Bad bet. Then Treasury Secretary Robert Rubin concluded that it wasn’t possible to save Russia without effectively turning Russian credit into US credit. He wasn’t willing to do that. He wasn’t willing to support the disbursement of the second tranche of Russia’s IMF program after Russian burned through the first tranche really quickly.

Not providing Russia more money then was a risky call. Kind of like letting Lehman fail. Russia, remember, had nukes. Lots of them. The national security types weren’t thrilled by the prospect of a bankrupt nuclear power.

Russia’s creditors (including Lehman) took large losses at the time. That presumably should have taught them a lesson or two about managing risk – it was more or less what I suspect Dr. Cowen would have prescribed.

It also implies that LTCM wasn’t the Lehman of 1998.

It was more like one of the institutions found to be swimming naked after Lehman defaulted.

Nor was LTCM the only big borrower that got a bit of help after Russia’s defautl.

Brazil, like Russia, had a lot of short-term debt that had to be rolled over. Now it so happens that most of Brazil’s domestic debt was owed to domestic banks not foreign investors – and that really helped. The analogy isn’t perfect. Brazil also had a pegged exchange rate. It, like Russia, had pegged to the dollar at too high a rate to be sustained after Asia’s crisis cut into global demand for commodities and reduced private capital flows.

Brazil not surprising came under a lot of pressure. But it also got a decent sum of money from the IMF. That loan supplemented Brazil’s reserves and allowed for a more orderly exit from its fixed exchange rate than otherwise would have been the case. They delay made possible by the IMF (and the government’s heavy intervention) in the foreign exchange market allowed a lot of Brazilian firms to hedge their dollar exposure, so they didn’t go bust when the real eventually was devalued. And Brazil didn’t default. Not in 98. Not in 99. And not in 2002, when it also had to draw on the IMF after Argentina’s default.

Ending moral hazard consequently would have implied letting Brazil go – not just letting LTCM go. The odds are that a Brazilian default soon after Russia’s default would have brought done a major financial institution or two, and brought about a major systemic crisis.

I am personally though glad that this wasn’t what happened. Brazil actually was suffering from a liquidity crisis as much as a solvency crisis. Or rather the IMF provided it with a cushion that allowed it to make the fiscal adjustment needed to assure its long-term solvency -- and that was something it was willing to do. That kept its liquidity crisis from morphing into a solvency crisis. The line between the two often isn’t as clean in practice as in theory (apologies for all the detail; I wrote an equation-free book on this with Dr. Doom before he was Dr. Doom).

I doubt Brazil would be better off today if it had defaulted in 98 or early 99. Defaulting on domestic government debt does bad things to the long-term health of any country’s domestic banking system. It creates a really bad hangover – and leaves a country permanently more vulnerable to a run.

Nor am I convinced that Dr. Cowen’s solution – standing aside as LTCM failed – would have ended moral hazard.

LTCM after all was an unregulated hedge fund. It wasn’t a regulated bank. Or a big – and sorta-regulated- broker-dealer.

If LTCM had failed, I suspect that policy makers would have stepped in to prevent any major regulated financial institutions from failing as a result of its exposure to LTCM, or its own LTCM-style bets. Rather than ending the expectation that big banks and big broker-dealers were too big-to-fail, the failure of LTCM might have reinforced that sense.

The real moral hazard in the financial system – in my view – comes not from expectations that if a firm like Goldman (or Lehman) makes a bad bet on a country like Russia (or a bad bet on US commercial real estate) the government will come in and protect the firm from losses on those investments. Rather it comes from the expectation on the part of those lending to places like Lehman and Goldman that these institutions are too big and too important to the financial system to fail, and thus it is safe to lend to them at low rates … no matter how leveraged they are or how many risky bets they are making.

If that is right, ending moral hazard in 1998 would have required allowing an institution like Lehman to fail. And, well, right now a lot of people think allowing an institution like Lehman to go bankrupt was a mistake.

To me the real failure during the last crisis was the failure of regulators to clamp down more seriously on leveraged institutions once the markets calmed.

Losses in Russia – and a close call with LTCM did lead to a bit more prudence for a while. Regulators did start to pay more attention to the financial firms that were providing a lot of credit to big hedge funds. But that started to seem a bit superfluous in a context where (for a period) the banks actually were lending less to hedge funds, in part because the big hedge funds were shrinking. Not just LTCM. Tiger too … even Soros.

And when the party got going again this decade -- and hedge funds and private equity firms and the broker-dealers and the banks (through off balance sheet vehicles) all started to gear up -- there was a team at the Treasury that wasn’t at all interested in regulating the financial sector. And the Fed – Greenspan especially – was never very keen on tight regulation.

Given all the scale of this year’s crisis, I certainly cannot rule out the possibility that Dr. Cowen is right and we would all be better off now if we had had a deeper crisis in 1998. A crisis that scarred the banks as deeply as it scarred most emerging markets might have produced a world where the banks wanted to increase their capital as badly as most emerging markets wanted to increase their reserves.

But I doubt that that outcome would have been possible without standing by and watching a lot more institutions than just LTCM fail. One big borrower -- Russia -- did fail rather spectacularly in 1998. Its failure created large losses for a lot of banks (far more than most were expecting, as some banks’ risk models at the time didn’t allow for a default on ruble denominated debt … ). The yen carry trade also unwound in ways that led to big losses at a lot of hedge funds in 1998. And that wasn’t enough.

A lot of the institutions that took lent large sums to Russia in 1998 were lending even larger sums to Russia in early 2008.

My bottom line: Forcing creditors to evaluate the real risk of lending to a large, highly leveraged financial institution -- i.e. the big banks and broker-dealers -- would have required a lot more that letting the market sort out LTCM without a gentle nudge from the Fed. It would have required allowing some solvent but illiquid institutions (and countries) to fail. That is a bit further than I would be willing to go.

*Bear Stearns excepted. Bear didn’t participate in the equity injection.

** Lending here should be read as shorthand for all credit exposure, even if it isn’t structured as a loan. And no doubt one complication of a plan based on “recapitalization from Lehman’s creditors” was that a lot of Lehman’s creditors had lent on a secured basis, and thus had little direct exposure to Lehman (though lots of exposure to a firesale of Lehman’s assets in the secondary market).

Relitigating 1998 …

Tyler Cowen argues that the “Committee to Save the World” made a mistake in 1998 by, well, saving the financial world. They thus missed an opportunity to teach the banks a lesson in sound risk-management.

He specifically argues that the Fed shouldn’t have called the big banks together to recapitalize LTCM. The recapitalization didn’t require any Treasury funds or draw on the Fed as a lender of last resort, so calling it a bailout obscured the meaning of a bailout – the fed catalyzed a private bailout but it didn’t do a true government bailout. But by acting, Cowen argues that the Fed set a precedent that creditors of big financial institutions don’t take losses, and thus encouraged bad bets.

I am not totally sure. The creditors of LTCM were the big banks, and they were in some sense “bailed-in.” To avoid taking losses on the credit that they extended to LTCM, they had to put equity into LTCM.*

It just so happened that the market recovered and it was possible for LTCM to exit many of its positions without taking large losses. The banks that took control of LTCM were able to unwind LTCM’s portfolio in a way that didn’t result in additional losses. But the result Cowen desired -- large losses for the banks and broker-dealers who provided credit to LTCM – was in the cards if LTCM’s assets weren’t sufficient to cover all its liabilities.

Lehman’s creditors didn’t get a chance to do a similar deal. There were too many of them -- and too little time. I suspect, though, that Lehman’s creditors and counterparties would be far better off if they had all agreed to pony up say 10% of the money they had lent to Lehman and in the process had provided Lehman with enough equity capital to allow it to be unwound in an orderly way.** They still would have taken losses, but those losses might well have been smaller – even counting the new money they put in – than Lehman’s creditors will incur as a result of Lehman’s bankruptcy filing.

Moreover, it seems a bit strange to look at LTCM in isolation.

LTCM, remember, came just after Russia defaulted.

And Russia was at the time considered the quintessential moral hazard play.

A host of financial institutions thought it was too nuclear to fail, and thus concluded that that they could safely pocket the high coupon on Russia’s GKOs (short-term Ruble denominated Russian securities) …

Bad bet. Then Treasury Secretary Robert Rubin concluded that it wasn’t possible to save Russia without effectively turning Russian credit into US credit. He wasn’t willing to do that. He wasn’t willing to support the disbursement of the second tranche of Russia’s IMF program after Russian burned through the first tranche really quickly.

Not providing Russia more money then was a risky call. Kind of like letting Lehman fail. Russia, remember, had nukes. The national security types weren’t thrilled by the prospect of a bankrupt nuclear power.

Russia’s creditors (including Lehman) took large losses at the time. That presumably should have taught them a lesson or two about managing risk – just as Cowen wanted.

It also implies that LTCM wasn’t the Lehman of 1998.

It was more like one of the institutions that was found to be swimming naked after Lehman failed.

The kind that are now truly getting bailed out. Many large financial institutions would be bust right now if not for Treasury capital injections and Fed liquidity support.

Nor was LTCM the only big borrower that got a bit of help after Russia didn’t get bailed out.

Brazil, like Russia, had a lot of short-term debt that had to be rolled over. Now it so happens that most of Brazil’s domestic debt was owed to domestic banks not foreign investors – and that really helped. The analogy isn’t perfect. Brazil also had a pegged exchange rate. It, like Russia, had pegged to the dollar at too high a rate to be sustained after Asia’s crisis cut into global demand for commodities and reduced private capital flows.

Brazil not surprisingly came under a lot of pressure. But it also got a decent sum of money from the IMF. That loan supplemented Brazil’s reserves and allowed for a more orderly exit from its fixed exchange rate than otherwise would have been the case. They delay made possible by the IMF (and the government’s heavy intervention) in the foreign exchange market allowed a lot of Brazilian firms to hedge their dollar exposure, so they didn’t go bust when the real eventually was devalued. And Brazil didn’t default. Not in 98. Not in 99. And not in 2002, when it also had to draw on the IMF after Argentina’s default.

Ending moral hazard consequently would have implied letting Brazil go – not just letting LTCM go. The odds are that a Brazilian default soon after Russia’s default would have brought down a major financial institution or two, and brought about a major systemic crisis.

I am glad that this wasn’t the course chosen at the time. Brazil actually was suffering from a liquidity crisis as much as a solvency crisis. Or rather the IMF provided it with a cushion that allowed it to make the fiscal adjustment needed to assure its long-term solvency -- and that was something it was willing to do. That kept its liquidity crisis from morphing into a solvency crisis. The line between the two often isn’t as clean in practice as in theory.

And I doubt Brazil would be better off today if it had defaulted in 98 or early 99. Defaulting on domestic government debt does bad things to the long-term health of any country’s domestic banking system. It creates a really bad hangover – and leaves a country permanently more vulnerable to a domestic run.

Nor am I convinced that Dr. Cowen’s solution – standing aside as LTCM failed – would have ended moral hazard among financial institutions.

LTCM after all was an unregulated hedge fund. It wasn’t a regulated bank. Or a big – and sorta-regulated- broker-dealer. If LTCM had failed, I suspect that policy makers would have stepped in to prevent any major regulated financial institutions from failing. Rather than ending the expectation that big banks and big broker-dealers were too big-to-fail, the failure of LTCM would have reinforced that sense.

The real moral hazard in the financial system – in my view – comes not from expectations that if a firm like Goldman (or Lehman) makes a bad bet on a country like Russia or a bad bet on US commercial real estate the government will come in and protect the firm from losses on those investments. Rather it comes from the expectation on the part of those lending to places like Lehman and Goldman that these institutions are too important to fail, and thus it is safe to lend to them at low rates … no matter how leveraged they are or how many risky bets they are making.

If that is right, ending moral hazard in 1998 would have required allowing an institution like Lehman to fail in a way that imposed large losses on Lehman’s creditors. Not just allowing LTCM to fail in a way that imposed large losses on firms like Lehman.

And, well, right now a lot of people seem to think allowing an institution like Lehman to go bankrupt in 2008 was a mistake.

To me the real failure during the last crisis was the failure of regulators to clamp down more seriously on leveraged institutions once the markets calmed.

Losses in Russia – and a close call with LTCM did lead to a bit more prudence for a while. Regulators did start to pay more attention to the financial firms that were providing a lot of credit to big hedge funds. But that started to seem a bit superfluous in a context where (for a period) the banks actually were lending less to hedge funds, in part because the big hedge funds were shrinking. Not just LTCM. Tiger too … even Soros.

Most macro funds got burnt on the yen carry trade in 98.

And when the party got going again this decade -- and hedge funds and private equity firms and the broker-dealers and the banks (through off balance sheet vehicles) all started to gear up -- there was a team at the Treasury that wasn’t at all interested in regulating the financial sector. And the Fed – Greenspan especially – was never very keen on tight regulation.

Given all the scale of this year’s crisis, I certainly cannot rule out the possibility that Dr. Cowen is right and we would all be better off now if we had had a deeper crisis in 1998. A crisis that scarred the banks as deeply as it scarred most emerging markets might have produced a world where the banks wanted to increase their capital as badly as most emerging markets wanted to increase their reserves.

But I doubt that that outcome would have been possible without standing by and watching a lot more institutions than just LTCM fail. One big borrower -- Russia -- did fail rather spectacularly in 1998. Its failure created large losses for a lot of banks (far more than most were expecting, as some banks’ risk models at the time didn’t allow for a default on ruble denominated debt … ). The yen carry trade also unwound in ways that led to big losses at a lot of hedge funds. And that wasn’t enough.

We shouldn’t forget: A lot of the institutions that lost a lot of money in Russia in 1998 were back lending huge sums to Russia in early 2008.

My bottom lines:

Getting rid of all moral hazard – and forcing creditors to evaluate the real risk of lending to a large, highly leveraged financial institution rather than bet that some large institutions were too big and too complex to fail – would have required a lot more that letting the market sort out LTCM without a gentle nudge from the Fed. It would have required allowing the kind of institutions that were lending to LTCM to have failed … and thus made it harder for the broker-dealers to expand their balance sheets and take on more credit risk in this decade.

And I suspect it would ultimately have meant allowing solvent but illiquid institutions (and countries) to fail. That is a bit further than I would be willing to go.

NOTE: I edited the post a bit to take out some repetition, though no doubt I could (and perhaps should) edit more ...

*Bear Stearns excepted. Bear didn’t participate in the equity injection.

** Lending here should be read as shorthand for all credit exposure, even if it isn’t structured as a loan. And no doubt one complication of a plan based on “recapitalization from Lehman’s creditors” was that a lot of Lehman’s creditors had lent on a secured basis, and thus had little direct exposure to Lehman (though lots of exposure to a firesale of Lehman’s assets in the secondary market).

More on:

Financial Markets

Close