PIMCO: The Honest Broker for the Week of October 10, 2014##
Joshua Brown: “Do we need to fire PIMCO?”: “In February of 2011, [Bill] Gross loudly proclaimed…
[that] Pimco Total Return had taken its allocation to US Treasury bonds down to zero. As recently as the previous December, Pimco Total Return had been carrying as much as 22 percent of its AUM in Treasurys…. Gross compounded the move by being extremely vocal about his rationale–he went so far as to call Treasury bonds a ‘robbery’ of investors given their ultra-low interest rates and the potential for inflation. He talked about the need for investors to ‘exorcise’ US bonds from their portfolios, as though the asset class itself was demonic. He called investors in Treasury bonds ‘frogs being cooked alive in a pot’. The rhetoric was every bit as bold as the fund’s positioning. It’s really hard to pound the table like this and then be flexible in the aftermath…
Yes, Bill Gross’s judgment in February 2011 that U.S. Treasuries were overbought has been an absolute disaster for PIMCO’s Total Return Fund vis-a-vis the market portfolio:
Holdings of ten-year U.S. Treasuries gained 20 cents on the dollar between Gross’s bet and the summer of 2012, as interest rates collapsed in the summer of 2011 and took another lurch downward in the spring of 2012. (They recouped 14 of those cents between the summer of 2012 and the end of the summer of 2013 “Taper Tantrum”, but today stand ten cents above their February 2011 value.
That being said, from Bill Gross’s perspective the belief that bonds as of February 2011 were overbought must have been irresistible, and not for reasons that were clearly wrong at the time. Since the start of 2008, 10-year Treasuries had been trading in a 2.5%-4% range appropriate for a safe asset in a low-inflation economy on the edge of or in the midst of a significant depression:
But at some point relatively soon, it seemed to Gross back in 2011, the economy had to recover to something like normal–in which case 10-year Treasuries ought to return to their 4%-5% trading range of the post-dot.com era:
if not to their 5%-7% trading range of the fast-growth 1990s:
And, Bill Gross thought, there was already light at the end of the tunnel: the economy was recovering, and the only things keeping expectations of recovery over the next five years from pushing up 10-year Treasury rates now was that the Federal Reserve was artificially restricting the supply of 10-year Treasuries via quantitative easing. And so when QE II ended, Gross was confident, Treasury rates would jump sharply–and Treasury bondholders would lose bigtime.
So what went wrong? Why did Gross’s expectations as of the winter of 2011 turn out to be so wrong? The standard answer is that long-term rates will not normalize until investors expect the normalization of short rates within half a decade, that short rates will not normalize until the Federal Reserve is confident that the zero lower bound crisis is over and will not return, and that that bond market confidence is further away now 3.5 years later than it was back in February 2011.
Paul Krugman is certainly the clearest exponent of this point of view–that it is all the immediate and obvious consequence of living in a liquidity-trap world:
Paul is certainly entitled to crow, crow some more, gloat, and crow yet again. He did write back before 2000 a book, The Return of Depression Economics, painting our current situation as not just a possible curious but even as a likely future scenario:
Paul Krugman: The Pimco Perplex: “Neil Irwin says…
A disastrous bet [Bill Gross] made against United States Treasury bonds in 2011 led to three years of underperformance and billions in withdrawals.
And Joshua Brown has some choice quotes:
Gross compounded the move by being extremely vocal about his rationale–he went so far as to call Treasury bonds a “robbery” of investors given their ultra-low interest rates and the potential for inflation. He talked about the need for investors to “exorcise” US bonds from their portfolios, as though the asset class itself was demonic. He called investors in Treasury bonds “frogs being cooked alive in a pot.”
But why was Gross betting so heavily against Treasuries? Brad DeLong tries to rationalize Gross’s behavior in terms of a coherent story about an impending U.S. recovery…. But Gross was… claiming that the Fed’s asset purchases… were holding rates down…. So why did he believe all that? It all comes down, I’d argue, to liquidity trap denial. Since 2008 the basic logic of the economic situation has been that the private sector is trying to run a huge surplus, and the public sector isn’t willing to run a corresponding deficit. The result is an economy awash in desired savings with nowhere to go. This in turn means that budget deficits aren’t competing with private borrowing, and therefore need not drive up interest rates. This isn’t hindsight; it’s what I and others have been saying since the very beginning.
But a lot of people… have just refused to accept this account. They have clung to the view that budget deficits must lead to higher interest rates. You might think the failure of higher rates to materialize, year after year, would cause them to reassess…. Instead… [they] made excuses… if only the Fed weren’t buying up the stuff. So QE2 acquired a much bigger role in their thinking than it deserved…. You can see why I found Gillian Tett’s apologia for Gross–that he was blindsided by central bank intervention–frustrating…. Gross of all people shouldn’t have fallen into this trap, since his own chief economist understood liquidity trap logic better than almost anyone. But finance people seem weirdly determined to believe in a macro canon whose hold on their perceptions appears to be completely unbreakable, no matter how much money it causes them to lose…
But is that really an adequate answer?
You can draw your standard IS-LM, with its four pieces:
- An LM Curve that shows the relationship between the safe nominal interest rate and the level of production as investors balance their money holdings they need to finance purchases against their bond holdings that yield them interest.
- An IS Curve that shows the relationship between the real risky interest rate and the level of production as households and businesses react to interest-rate incentives to buy more or less and so keep factories running or idle.
- The inflation wedge: the difference between nominal and real interest rates.
- The risk-premium wedge: the difference between risky and safe interest rates.
It looks like this in the liquidity trap, at the zero nominal interest rate lower bound at which short-term bonds and cash money become perfect substitutes:
In the liquidity trap expansionary monetary policy–the Federal Reserve’s buying bonds for cash and so pushing the LM Curve out and to the right does not do anything to the economy’s equilibrium (unless it has knock-on effects that increase expected inflation or diminish the risk premium):
But expansionary fiscal policy–borrowing money and buying stuff–increases output and employment without raising interest rates as long as monetary policy is sufficiently accomodative:
But you can make Gross-like arguments in this diagram. If you forecast that rapid normalization will both reduce the risk premium and shift the IS Curve out as consumption and investment spending return to normal plus see government debt issue as increasing expected inflation, a liquidity trap now is perfectly consistent with high equilibrium nominal interest rates soon, and thus with long-term interest rates now that ought to be forward-looking, hence quickly normalizing:
From this perspective, Paul Krugman’s 2011 declarations that we are in a liquidity trap now hence more deficits now will spur recovery without raising interest rates appeared to beside the point. Yes, that was true now. But in Bill Gross’s world, spending and thus the IS Curve was going to normalize, this risk premium was going to normalize, and the flood of government debt issues were going to raise expected inflation. That all of these were going to happen meant that forward-looking long-term nominal interest rates should rise by a lot soon. That was Bill Gross’s bet. And it went wrong.
So why? Why didn’t risk premiums fall? Why didn’t business and consumer spending normalize? Why didn’t debt issue produce higher actual and expectations of inflation? I think we need a better answer than Krugman provides…
As I see it, Bill Gross made four analytical mistakes in the winter of 2011:
-
He had much too much confidence in the market economy’s ability to stabilize the macroeconomy itself–or perhaps the government’s desire and ability to stabilize the macroeconomy.
-
He draw an inappropriate parallel between the actions of individual investment banks that push asset prices away from fundamentals and the actions of a central bank.
-
He failed to properly understand the link between the diminished relative risk tolerance of bond investors and the price of Treasury bonds.
-
He failed to remember the Dornbusch Rule.
Let me run through all of these in turn:
(1) The Confidence Fairy
Bill Gross’s first mistake, I think, is that he had been visited by the Confidence Fairy.
Look at the United States since 1948. Odds are, if the unemployment rate is above and employment and production are below their trend values, all three will get two-thirds of the way back to its trend value within two years. To bet on the American economy staying in the bust is almost surely to lose–at least, it is almost surely to lose until 2008 and thereafter. We can argue over whether the pre-2008 American economy possessed strong self-regulating equilibrating forces in its private-sector core or whether it was clever and effective monetary policy by the Federal Reserve that America’s pre-2008 busts short and its inflationary-spiral booms–with the exception of the 1970s–short as well.
What is important to note is that, since 2008, this time things are different. Two years after the business-cycle trough of October 2009 the unemployment rate was not 2/3 but only 3/10 back to the previous normal. The recovery of output back to trend was even less: not 3/10 but 1/7. And the recovery of employment as a share of the adult population? Nowheresville.
If you expected a normal speed of recovery since 2009 and a normal speed of reversion of the Federal Reserve to normal interest rate policies, and if you bet on that speed of normalization, you lost your shirt–as PIMCO Total Return did. That said, this was an elementary error that Gross should have avoided: there was a great deal of evidence available as of early 2011 that this time was different as far as the speed of recovery was concerned.
(2) Risk Tolerance and Yield Spreads
Gross’s third mistake, I believe, was failing to understand how 2008-2009 were likely to change the relationship among asset yield spreads.
The natural benchmark for a risk-free real return is the current short-run Treasury note rate minus the current inflation rate. What is the natural corresponding benchmark for a risky return? I have always taken it to be the permanent earnings yield on a broad stock market portfolio:
The real yields on bonds more risky than Treasuries and less risky than diversified common-stock equity investments move in the field of force defined by these two poles–the Treasury (real) rate and the common-stock yield.
It was conventional wisdom before 2008 that the spread between equity and Treasury yields was shrinking as a result of improvements in the desire and ability of investors to properly structure the risks that their portfolios were to bear–and the conventional wisdom after 2009 was that the financial crisis had not permanently deranged previous patterns. Gross was, therefore, looking forward to a near- and medium-term future in which equity yields would be on the order of 5%/year, inflation would be on the order of 2% to 3%/year, and the premium yield of equities relative to 5- to 10-year Treasuries would be on the order of 2% to 3%/year. That would mean that the 10-year Treasury annual yield would be attracted to something more than 5%.
- He failed to properly understand the link between the diminished relative risk tolerance of bond investors and the price of Treasury bonds.
(3) The Dornbusch Rule
Bill Gross’s third mistake, I believe, was to neglect the Dornbusch Rule. The Dornbusch rule is that your calculations of fundamentals may be accurate, and you may have high confidence in them, but nothing requires that the market has to have high confidence in your beliefs about fundamentals. Thus markets can remain far away from equilibrium for far longer than you, who understand and are dazzled by your analytical insights, think possible. As the spouse of one senior hedge-fund official put it: your theory of the world is that the market is inefficient when you put a trade on but will rapidly become efficient thereafter–where “rapidly” means “before your clients lose their patience with you”. Bill Gross forgot that. And that is one reason that he put his bet on not with trepidation and with statements about how the balance of risks suggested underweighting Treasuries, but rather with the extravagant rhetoric that left his reputation hostage to the trade turning out well.
(4) The Washington Super-Whale
I wrote about this before, in a similar context, a year and a half ago. Bill Gross and those who thought like him faced a world in the aftermath of the financial crisis in which (a) they expected a return to normal levels, (b) they were confident that modern financial engineering would drive a return to small spreads, and (c) they believed the market rational enough to in short order adopt their view of fundamentals and push asset prices to the fundamental configuration. Yet, as of early 2011, it had not happened. Why not?
The natural answer if you are a financier to why prices stubbornly refuse to return to fundamentals is that one of two things are happening: (a) sentiment has gone mad and there is a–positive or negative–bubble; or (b) some huge trader is taking on a very risky and almost surely losing position because of the price pressure the enormous size of their bet is creating. And in the aftermath of the financial crisis, it seemed obvious that it was (b), and that the enormous trader was the Federal Reserve.
We saw what Bill Gross and company thought was going on in 2011 play out in miniature, with JPMC in the role in which they had cast the Federal Reserve, the following year. In February 2012, traders noted one particular underpriced index–CDX IG 9. There was an obvious short-term moneymaking opportunity: Buy the index, sell its component short, in short order either the index will rise or the components will fall. But April rolled in, and the gap between the price of CDX IG 9 and what the traders thought it should be grew. Their bosses asked them questions: “Shouldn’t this trade have converged by now?” “Have you missed something?” “How much longer do you want to tie up our risk-bearing capacity here?” “Isn’t it time to liquidate–albeit at a loss?” So the traders began asking who their counterparty was, and found he was singular–“the London Whale”. So they got annoyed. And they went public, hoping that they could induce the bosses of the London Whale to force him to unwind his–very risky–position. And so we had “‘London Whale’ Rattles Debt Market” and similar stories.
The London Whale was Bruno Iksil. His boss, Ina Drew, took a look at his positions and found that JPMC had a choice: They could hold CDX IG 9 until maturity while singing “Luck, Be a Lady Tonight!” and bet JPMC on a single crapshoot–make a fortune if a fewer-than-expected number of the index’s 125 companies went bankrupt and lose JPMC to bankruptcy if more did–or they could eat a $6 billion loss and go home. Since JPMC could not survive in the absence of an unlikely government bailout if its net worth went negative even for a day, Drew stood Iksil down and the traders had their happy ending.
What Bruno Iksil did was what Bill Gross–and Cliff Asness, and a huge host of other Wall Streeters–thought that Ben Bernanke was doing. He had run the Federal Reserve’s balance sheet up to absurd proportions, and in so doing had pushed interest rates well below and Treasury and MBS bond prices well above fundamentals. Eventually, and sooner rather than later, Gross and Asness and company thought, the Federal Reserve would have to revert and unwind its grossly overleveraged position–and when it did so it and everybody else long Treasury bonds would lose a fortune, and those smart enough to bet on fundamentals by shorting Treasury bonds would profit.
Bruno Iksil had been pulled up short by his boss Ina Drew’s unwillingness to hold his positions to maturity and so risk JPMC’s bankruptcy. Ben Bernanke, they thought, ought to have been pulled up short by his unwillingness to risk an inflationary spiral–for Bernanke had financed his Treasury bond purchases by issuing reserve balances, and when banks became confident enough to diminish their excess reserves by using them to back deposits and when businesses and households became confident enough to spend those deposits, then there would be a devastating inflationary spiral unless the Federal Reserve had previously unwound its position. It was, they appear to have thought, unprofessional for Ben Bernanke not to have already unwound his positions as of the winter of 2011 and for him to in fact be adding to them via QE II. Thus, I think, one purpose of Bill Gross’s bet and declaration was to do to Ben Bernanke what the following year’s leaks about ‘the London Whale’ did to Bruno Iksil: force recognition of the extraordinary risks Bernanke was running, and so start the unwinding of the Federal Reserve’s balance sheet, and so trigger the return of bond prices and interest rates to fundamentals.
The problem, of course, is that the parallel is faulty. If JPMC has a moment of negative net worth, it is toast. If JPMC’s positions have pushed asset prices far enough away from fundamentals that it is perceived to in the future have a significant chance of giving it a negative net worth, it is also toast: the fact that the first people to unwind their positions vis-a-vis JPMC get out whole while those who do not may not triggers a shadow bank run on JPMC’s non-inertial liabilities, and in order to cope with that run JPMC has to liquidate assets at fire-sale prices and that triggers the negative net worth that was feared as a future possibility. There is thus a very sharp and direct chain of actions with very hard incentives at every stage leading from too-much leverage to catastrophe.
By contrast, if the Federal Reserve’s purchases are perceived as having pushed asset prices away from fundamentals–or, rather, “fundamentals”–well, then what? Some people sell Treasuries and buy other assets–equities, say–in an attempt to profit from the forthcoming return of interest rates to “fundamentals”, and then what? Others have the cash that the Fed issued to finance its born purchases in their pockets, and they wonder if they should spend it, and then what? Well, nothing–except to the extent that businesses with higher stock prices decide that they can afford to cut back on share repurchases and employ more people to build capacity instead, and except to the extent that those with the extra cash don’t just wonder but actually increase their spending. And if they do? Then the economy recovers. Strong market reactions to Bruno Iksil’s position would be the source of catastrophe for JPMC; strong market reactions to the Federal Reserve’s QE policies would–if they were to occur–be the desired effect of the Federal Reserve’s policies.
But doesn’t the Federal Reserve have to unwind its balance sheet? After the economy has recovered to a normal level of activity, yes. But won’t the Federal Reserve lose a fortune when it does so? Yes–but so what? The Federal Reserve does not have real shareholders, and does not have a real fiduciary duty of wealth maximization to the fake shareholders it has. If when the Federal Reserve has finished unwinding its position in order to keep inflation from accelerating and finds that it has no assets and $1 trillion of liabilities in the form of Federal Reserve notes and reserve deposits outstanding, so what? The reserve deposits are valuable because they allow you to accept commercial bank deposits. The Federal Reserve notes are valuable because you can pay your taxes with them. It’s not the same with the liabilities of JPMC, which are of very dubious value if JPMC is or is feared to possibly in the future be bankrupt, for which the lack of secure fundamental value as JPMC approaches bankruptcy creates the possibility of terrifyingly rapid and destructive bank runs. Demand for a private bank’s liabilities can collapse quickly and suddenly in a bank run. Demand for a central bank’s liabilities cannot–especially when the central bank’s liabilities are a reserve currency.
It was, I think, this faulty analogy between a private investment bank that has over-expanded its balance sheet by purchasing some asset class on a large enough scale and so pushed prices away from fundamentals and the Federal Reserve’s balance sheet expansion that, I think, played a key role in Bill Gross’s analytical error. But that was just the final miscalculation. Before it came (a) misjudgments about the speed with which normalcy would be reattained, (b) misjudgments about what risk yield spreads would be when and if normalcy were to return, and (c) misjudgments as to how long the market could stay out-of-equilibrium–longer, as John Maynard Keynes does not appear to have said, than Bill Gross could remain if not solvent at least in charge of PIMCO Total Return.
3907 words
