Sunday, January 30, 2011

Black Swan Events and How to Hedge Them

from Zero Hedge:

With all the hoopla over Egypt some have forgotten that this is merely a geopolitical event (one of those that absolutely nobody, with a few exceptions, was talking about less a month ago, so in many ways this is a mainstream media black swan which once again exposes the entire punditry for the pseudo-sophist hacks they are), and that the actual mines embedded within the financial system continue to float just below the surface. Below we present the five key fat tail concerns that keep SocGen strategist Dylan Grice up at night, which happen to be: i) long-term deflation, ii) a bond market blow-up, iii) a Chinese hard-landing, iv) an inflation pick-up, and v) an Emerging Markets bubble. Far more importantly, Grice provides the most comprehensive basket of trades to put on as a hedge against all five of these, while also pocketing a premium associated with simple market beta in a world in which the Central Banks continue to successfully defy gravity and economic cycles. For all those who continue to trade as brainless lemmings, seeking comfort in numbers, no matter how wrong the "numbers" of the groupthink herd are, we urge you to establish at least some of the recommended trades in advance of what will inevitably be a greater crash than anything the markets experienced during the depths of the 2008 near-cataclysm.
But before we get into the meat of the piece, we were delighted to find that Zero Hedge is not the only entity that believes that providing traditional annual forward looking forecasts is nothing more than an exercise in vanity (and more often than usual, error).

At this time of the year we’re supposed to give our predictions for what’s in store for the year ahead. The problem is I don’t have any. Not because making forecasts is difficult. It isn’t. It’s just pointless. Instead, I suggest getting in touch with our inner Kevin Keegan, the hapless former England football manager who, facing the sack after a bad run of results famously lamented “I know what’s around the corner, I just don’t know where the corner is.” The more people construct portfolios on the assumption that they can see the future, the greater the opportunity for those building portfolios which are robust to the reality that we can’t.
That said, no matter how ridiculous the act of Oracular vanity ends up being, those who charge an arm and a leg for their "financial services" continue to do it, only to be among the first carted out head first when reality is imposed upon them and their blind belief that this time is different and the crowd is actually right. Few are willing to accept and recognize the humility that they really know little if anything about how a non-linear, chaotic system evolves. Which is once again why we believe that Grice is among the best strategists out there: in his attempt to hedge the stupidity of the crowd, he has coined a term that may well be the term that defines 21st century finance and economics: instead of foresight, Grice believes the far more correct term to explain the process of prognostication should be one based on foreblindness.
In financial markets, craziness creates opportunity. It affects only prices, not values. And one of the craziest afflictions I know of is our faith in our ability to see the future. Indeed, there isn't even an appropriate opposite to the word "foresight"  in the English language. So I'?m going to make one up. And rather than build a portfolio based on the pretense we have foresight, let's explore some ideas for building one that is robust to our foreblindness.
This is the kind of insight that one will never find from a TBTF "strategist"... And one wonders where all those softdollars go.
So now that we know that unlike the traditional cadre of sell side idiots who are always wrong in the long-run, Grice actually admits that he has no clue what will happen, which is precisely the reason to listen far more carefully to what he has to say.
Let's dig in:
Here are some things I think are true:
  • developed economy governments are insolvent
  • Japan is the highest risk developed market (DM) to an inflation crisis (though it might be Greece)
  • there is too much debt around
  • China?s economic model is biased towards misallocating resources
  • every country which has industrialized has experienced nasty bumps on the way
  • China and the US are in the early stages of an arms race
  • demographic trends suggest more conflict in the oil rich regions of the world
  • bottlenecks are developing in key commodity markets
  • the only thing central banks are good at is blowing the bubbles that cause the crashes which are used to justify their existence
  • market prices only reflect fair value by accident and in passing
  • most people don't think these things are important
  • they might be right.
Here are some things I know are true:
  • perceived uncertainty causes emotional discomfort which isn't conducive to good decision making
  • all the above situations have the potential to cause significant asset price volatility
  • I have no idea when.
What to do? To my mind, the ideal is not to make huge bets on particular events happening because failure of the expected event to materialize will materially endanger your capital. Instead, the ideal is purchase insurance at a price which won't materially pressure the returns from your core portfolio of investments if the event fail to materialize, but will protect capital from significant impairment if it does.

Is such an ideal attainable? By evaluating insurance and using the same valuation discipline you'd apply to anything else, I think it is. So what follows is not a list of recommendations here, or even any suggestions. Everyone should do their own homework. What follows is an illustration of why I think the macro research we've been doing is relevant and can be used to lower portfolio risk. The insurable risks I'm most worried about at the moment are:
  • long-term deflation
  • a bond market blow-up
  • a Chinese hard-landing
  • an inflation pick-up
  • an EM bubble
The first thing you?'ll notice is that these aren't all consistent with one another. It's difficult to get a Chinese hard landing and an EM bubble at the same time, for example. But internal consistency is overrated. It's only relevant for point-in-time forecasts, and the assumption underlying this entire exercise is that I haven't a clue if/when any of what follows is going to happen. At the risk of repetition, I'm interested in the possibility of building a profitable portfolio which is robust to my ignorance.
Let's take a look at the five fat tails in detail:
Long-term deflation
Not surprisingly, Grice gives the least amount of weight to the one thing most troubling to such economic disgraces as Ben Bernanke and Paul Krugman. Yet it should not be avoided. After all there are many deflationists out there, who believe that the Fed, which has now clearly telegraphed it is all in on reflating (or after the Fed, the monetary collapse deluge) may actually succumb to what has been ailing Japan for two decades.
According to economists the primary risk faced by economies is that a huge deleveraging spiral becomes self-fulfilling: deleveraging reduces demand, which lowers prices, which further lowers demand, and so on. The idea was first developed by Irvine Fisher in the 1930s to describe the great depression, and has been used to explain the ?First Great Depression? of the 1870s and Japan since the early 1990s.

Paul Krugman says everything has changed because we?re in a liquidity trap. The fear of prolonged deflation is what keeps poor old Ben Bernanke awake at night. And maybe that's the clue. At our London conference this year, James Montier said that Bernanke  as the worst economist of all time. Now, I'm not sure I agree with James on this one because I can't make up my mind, sometimes I think it's the Bernanke, other times I think it's the Krugman. But usually I think nearly all economists to be the joint worst economists of all time. So I have a lot of sympathy with the idea that if the consensus macroeconomic opinion is worried about something, it probably isn't worth worrying about. In fact, if they worry about deflation, I'm going to worry about inflation.
We couldn't have said it better ourselves.
So how does one trade deflation insurance?
More importantly though, deflation insurance is expensive. The following chart shows the price of 5y 0% US CPI floors to be trading for just under 200bps. The way these floors work is that they provide the owner of the contract with the right to payments equal to the rate of deflation. Since the floors in the chart have a five-year maturity, they entitle the owner to five annual payments. For example, if inflation was -1% in year one, the owner would receive 100bps of the notional value of the contract. If inflation was -1% in year two, he'd receive another 100bps. And if the rate of deflation remained at -1% for years three, four and five, he'd receive 100bp cash flow for each of those annual payments so that over the life of the contract he'd have received a total cash flow of 500bps. So if you're worried by the prospect of CPI deflation, this is the product for you.

And skeptical though I am of the debt deflation hypothesis, Western demographics worry me. Although we don?t know what aging economies look like, we know that the glimpse into the future provided by Japan isn't encouraging. So I do take the scenario seriously and would be happy to put the hedge on at the right price. The problem is, I don't think the price is right. I think this insurance should be sold, not bought.
Chinese Hard Landing
While Grice is obviously far less worried about a systemic deflation scenario arising out of events in the US, what may happen in China is obviously a far riskier proposition, and one that could generate deflation out of the proverbial "Hard Landing." Luckily there is an instrument with some wonderfully convex properties to hedge this...
Albert calls China a ‘freak economy.’ Certainly, running with an investment to GDP ratio of over 50% doesn't seem normal. Neither does keeping interest rates at 5% when the economy is growing by 15% in nominal terms each year. Such lax monetary conditions have helped land prices rise by 800% in the last seven years, according to NBER economists. And when you come to think of it, more recent examples of real estate inflation fueled by negative real interest rates: Ireland, Spain, the US didn't end too well. Jim Chanos says China is a shortseller's dream and that there's not one company he's looked at that passes the accounting sniff test. And if Jim Chanos, who built a very successful business around spotting accounting gimmickry says something like that, my guess is he's right.

Taking a step back though, as far as I?m aware all industrialized countries have experienced financial crashes. It seems a part of the maturing process. Why should China be any different? A credit crisis wouldn't necessarily mean the end of the China story any more than the panic of 1873 meant the end of America's, (though US demographic prospects were considerably more favorable at the end of the 18th century than China's are today). For the record, I think the Chinese have a bright future. My son is learning Mandarin. But when I look at the numbers I can't help but think there's going to be a crash and that it's going to be  quite unpleasant. It's just that my guess as to when it's going to happen is as good as Kevin Keegan's.
What happens if and when the inevitable crash happens? One word - Australia. Another word(s): 10x-20x payout.
When it does happen though, the Australian economy will be toast and its government bond yields will collapse. During the panic of 2008, AUD 10y swap rates fell around 3% to 4.40%.

The panic of 2008 was a "good crisis" for Australia though. A Chinese crash would be more serious.

And you can get pretty attractive odds on AUD rates collapsing. The following chart shows the payout available using AUD receiver swaptions prices with a three-year maturity, based on the 10y swap rate. Effectively, these are put options that pay out when rates fall below the strike price. The prices I've used here are from Bloomberg based on the swaptions striking at about 5.5% (i.e. 100bp below the current rate of 6.50%). What's interesting is that at current prices, if Australian swaps were to break their 2008 lows, you'd be making about 10x your premium (for the record, these swaptions are priced at about 120bps, or 40bps per year over three years, which is about the same as the annualized revenue you'd get if you sold the CPI floors discussed above). If swap rates fell by 300bps as they did during the panic of 2008 the rate would fall to 3.5% and you'd make nearly 15x your premium. To repeat the point I made earlier, this isn't a recommendation. It's just a starting point (my guess is that you'd find more attractive payouts as you went further out of the money with the strike price, and that  capital structures of Australian banks, property companies and levered resource stocks would be worth looking into too).
Asset Bubbles
Grice provides one of the best and most succinct explanations of bubble mentality we have read to date:
For reasons I won't go into now, but which are probably obvious from what I just wrote about China, I think EM is riskier than it seems. I'm not even sure I feel comfortable valuing EMs yet. So should EMs bubble up, the risk for investors sharing my concern is that they'll be faced with quite a nasty dilemma: do they buy something they don't feel sure is cheap because everyone else is and they're scared of underperforming, or do they stick to their principles and prepare themselves to take on the business and career risk of underperforming their competitors, seeing clients withdraw their funds, and possibly finding themselves out of work?

And the sad reality is that ultimately nearly everyone gets hurt during a bubble. Sceptics get hurt as it inflates, believers get hurt when it bursts. George Soros says when he sees bubbles he buys them. He?'s been pretty good at selling them at the right time too. But most of us aren't so clever.
Regardless of the psychology behind each and every bubble, the good thing is that there is a good way to hedge this risk outright.
One way to hedge the inflation of a bubble, rather than its bursting, is to buy out of the money call options on the equity indices. Calls are usually cheaper than puts ? I think because fear is a more powerful emotion than greed and the tails in equity markets tend to be on the downside. But the following chart shows that that difference (or skew, the difference in implied volatilities between puts and calls) is close to unprecedented highs, at around 4.5 vol points (the chart shows skew for the S&P500 though other equity indices show a similar picture).

In other words, the upside is close to unprecedentedly cheap relative to the downside. If you could get two year call options 30-35% out of the money for 130bps per year you?'d be getting good value (of course you could make this zero cost, or even ve cost by selling puts to fund the purchase, and you could do it in such a way that your downside risk would be similar to that of holding stock, but I'm no derivatives strategist- as usual, if you want to talk about this stuff to people who know more than I do, speak to your SG derivatives salesman, or ask me and I?'ll put you in touch).
Hyperinflation
A topic near and dear to many. Luckily, once again, one which can be hedged proximally in a form that generates massive returns should it transpire there where it most needed: no, not the US. Japan. In fact, if Grice is right about Japan, his proposed trade takes the returns generated by Paulson in shorting subprime... And magnifies them by about a million.
Historically, bankrupt governments have used inflation to alleviate their indebtedness. I doubt things will ultimately be different this time. And as regular readers know, I think Japan is the country closest to the edge. All DM governments have the same problems: they've made promises to their electorates which they're unlikely to be able to keep. But while there's time for European and US governments to fix the problem, for Japan I think it's already too late. John Mauldin says the Japanese government debt position is a “bug in search of a windshield”. I agree with him.

I've already written too much this week, so I don't want to rehash all the stuff I've already written on Japan and which regular readers will be familiar with. But if you chart past episodes of extreme inflations with how stock markets behaved during the episodes, you invariably find something similar to what happened to Israel in the 1980s.

In Steven Drobny's excellent “Hedge Funds Off the Record” (which I consider a must read - almost every interview oozes with profound risk-management wisdom), Steve Leitner talks about buying out of the money call options to hedge against such a hyperinflation. Buying 40,000 strike Nikkei calls with a ten- year maturity, with a payout in a strong currency can be done for around 40bps per year. And to give you an idea of how explosive that asymmetry might be, if Japan was to follow the Israeli experience from here, the Nikkei- currently 10,500 would trade at around 60,000,000 (sixty million). So putting even one-tenth of your notional into that kind of hedge would cost 4bp per year (for reference, the Nikkei currently offers in excess of a 2% annual yield, while some JREITS offer in excess of 4% - I'd argue that 40bp is a bearable burden, and 4bps certainly is).
Bond Market Blow-up
When a few weeks ago we presented Sean Corrigan's chart which we dubbed the Great Regime Change, few put two and two together, and realized the vast trading implications of this chart. And they are profound. As Grice rightfully observes, they stand at the base of nothing less than the hedge against that most critical of fat-tail events: a bond market blow up, one which is getting increasingly more probable with every single $X0 billion UST bond auctions (the bulk of which is now monetized directly by the Fed).
One obvious way to hedge against a bond market blow-up is to use the swaptions market as we did in the Australian market to hedge a Chinese crash, only this time buying payer swaps, which are effectively call options on rates. But I thought I'd show you something I think is a bit more interesting: the correlation between the S&P500 and bond yields.

Bonds represent poor value in my opinion, with little margin of safety to protect against the very real risk that governments try to inflate away their debts. But one good reason to continue holding them is that they protect risk asset positions during the '?tails'?. The following chart shows that over the last ten years the correlation has been volatile, but positive: when equities have fallen so have bond yields, offsetting losses in the equity portfolio as bonds benefit during "risk-off" events.

When inflation expectations were (probably) around zero (before the 1960s) the correlation between bonds and equities was zero too. But look what happened during the 1960s when inflation expectations broke (this was during the Vietnam war, as the Bretton-Woods system was coming under pressure and as the bear market in bonds was getting into full swing). The chart shows that the correlation went negative. When bond yields rose equities fell because government bonds were reflecting the same tensions that were pulling down equity valuations (fear of ever-higher inflation).

As the bond bear market reached its climax in the early 1980s, the correlation remained negative. But as the worm turned, and central banks across the developed world made new and credible commitments to stop printing money, a bond market rally was born. And as inflation expectations began to fall, what was good for the bond market was good for the equity market. Now, falling yields coincided with rising equity prices and so the correlation remained negative. But during the last ten years, inflation expectations have been roughly stable and, if anything, slightly biased towards the deflationary side. So what?s been good for bonds hasn't been good for equities, and the correlation between yields and equity prices has been positive to reflect that.

The point is this: if governments are insolvent, and the government bond market becomes a source of risk once again (as opposed to the nonsensical "risk-free" description it has somehow obtained in recent years) what's bad for the bond market will be bad for risk assets too. As yields rise, risk assets will fall. The correlation will go negative. Bonds will provide less protection against the tail events than they have done in recent years because they will be a source of the tail event.
For all those who figured this out based on the Corrigan chart, congratulations. This could well be the holy grail of the biggest black swan insurance trade of all. For those who haven't quite grasped it, here is some more from Grice:
This correlation is tradeable. Any bank with a derivatives operation must have an implicit correlation exposure between products they've sold options on. So for derivatives houses, correlation is a by-product in much the same way that molybdenum is a by-product of copper miners. and correlations like this trade in the IDM market. And sometimes that means you can get it for a very good price. I recently heard of a correlation trade between the S&P500 and the US 10y swap rate done at 40 correlation points, which seems a decent enough price to me (of course, selling at 50 points would give you even more margin of safety), although current pricing is at around 30 I believe. Pricing can be volatile though and waiting to sell in the 40-50 range seems sensible to me. It would hedge risk positions against a regime in which government bonds were seen as the source of risk, rather than the reliever of it.
Finalizing the Black Swan Insurance basket.
Let's add it all up and see how much it would cost to insure our portfolio. If we were to sell the 5y US CPI floors for 200bps (40bps annualized); buy the 3yr AUD receiver swaption for 120bps (-40bps annualized); buy 2yr 30% S&P500 calls for around 130bps (-65bps annualized) and bought one tenth of our notional on NKY calls for 40bps (-4bps annualised) the net upfront cost would be 90bps (200bps-120bps-130bps-40bps). If we wanted to hedge the risk of bond market turbulence with a correlation product, this would cost nothing upfront because it would be done on a swap basis with the bank. On a roughly annualized basis our cost would be 69bps each year.

Of course, we'd have a maturity mismatch because our hedges would have different time horizons. So we'd have to adjust them from year to year. We'd also be more vulnerable to deflation because we don't think the deflationary hedges offer value. So our portfolio wouldn't quite be bullet-proof because it would be tilted towards inflationary outcomes. But we'd have insurance against deflation with the Australian receiver swaption. And since the correlation swap hedges us against any bond market blow-up which also blows up the equity market, we can feel more comfortable allocating some capital towards bonds we think might offer good value (not that there are many, I'd say maybe about 20-30% in Australian and New Zealand bonds).

I'd put 10% in gold. I'll explain more in another note but for now, although I've said I'm not a fan of plain commodities as investment vehicles because buying commodities was equivalent to selling human ingenuity, I exclude gold from that logic. I prefer to see buying gold as buying into the stupidity of governments, policy-makers and economists, and I'm comfortable doing that.

With the exception of Japan (which we'd be hedged against anyway) I'm not so worried about "traditional" CPI inflation any time soon. At the moment, I think the first signpost on the way to that kind of crisis will be via the bond market, which the correlation swap should protect us against. That and my gold holdings would make me comfortable allocating 20%-25% cash. I still think risk assets are generally overvalued and cash is the simplest insurance "option", whose relative value rises proportionate to the decline in other assets. So let's say I'm 20% in cash,10% in gold, and 20% in mainly Australian and New Zealand government bonds. That leaves just under 50% of my capital for me to put into the equity market (the 69bps per year for my insurance bucket to be precise).

Which equities? I've always thought investing in index funds to be crazy, but nearly everyone does it and it's a part of the craziness we can use to our advantage. The EMH says that market prices are always broadly efficient because all market participants respond to all available information. But around 10% of the market is explicitly passive and probably another 50% is benchmarked and therefore implicitly passive. In other words, the overriding variable for the majority of equity investors is a company's weight in the index! Intuitively therefore, the prices can't fairly reflect fundamental value, which means that at any point in time, there will be lots of stocks which are mispriced.

The following chart shows two lines. The red line shows the cumulative return to buying stocks in the cheapest decile, while shorting stocks in the most expensive decile (I define value as the discount relative to the estimated intrinsic value - a methodology I've been meaning to write up in detail for several months now but which I will definitely do within the next few weeks). Using a monthly rebalance, the annualized return is 750bps. This shows that there is meaningful alpha in identifying and owning those stocks trading at a discount to intrinsic value. The black line shows the relative outperformance of the top decile against our wider stock universe. (In passing, note that this value strategy underperformed in the late 1990s during the tech bubble, and remember that this is the reason our hypothetical portfolio has out of the money call options.)

The relative outperformance of this long-only basket has been 330bps. If I expect a stock market return of 5% per year over the coming years, that 330bps outperformance is highly significant. It means we only need to put 60% of our capital into that basket of stocks to generate the same incremental return as a market portfolio would generate. So owning 50% isn't as cautious as it sounds.
The bottom line, and the reason why we think this is a great basket trade, is that it makes money in a normal environment while at the same time, providing great positional hedges to those 5 events which sooner or later are bound to happen.
In the sort of world in which everything is normally distributed, well behaved, and in which our insurance expires worthless (i.e. the sort of world most economists forecast), we'd still be making decent returns. And while there's no such thing as a truly bullet proof portfolio, we'd have done so with far less embedded risk. Because if any of the scenarios I've explored here come to pass we'd be in a much better position to take advantage of the distressed selling of others.
That said, and as Dylan would be the first to acknowledge, if and when everyone is positioned with precisely this hedge on their books, it will be something totally different that will cause the next great financial wipeout. But until then, those who step in first, will benefit from appreciating prices precisely on their hedges. At that point it will be up to the principal to decide whether to take profits or to hold off until the bitter end. The problem, however, at least the way we see it, is that should any of these five black swans occur, any currency that one generates as a result of a successful trade, no matter the P&I, will probably not be all that useful for the world that materializes at T+1.