The Bubble Puzzle

For all these people who think the rise in oil prices is due to a bubble -- and I'm not saying it isn't, 'cause I honestly don't know -- shouldn't they put their money where their mouths are and sell oil futures short? I mean, if the high prices are due to speculation and not fundamentals, and if you think you know the fundamentals better than the speculators, shouldn't you be able to beat them at their own game and get rich in the process?

Am I missing something here is or is all this talk of bubbles just cheap talk?

Share this

The dynamics of a bubble

I don't know if it's a bubble or not, but it's not simple to play a bubble short.

Remember that bubble behavior is parabolic: not only does the price keep increasing, but rather, the price keeps increasing faster and faster. This leads to a blowoff top (like the NASDAQ had in 2000).

An argument can be made that the rational stragegy would be to play a bubble long. When your money doubles faster and faster, it makes sense to try to exploit it. Sure, you may end up being the "greatest fool", but it's more likely that you're just going to sell it at much higher prices to a greater fool. In less than 2 years, the NASDAQ nearly quadrupled (1400 --> 5200), and certain stocks went up by an even greater amount. Because this strategy is rational (IMO), it is what drives a bubble. A bubble wouldn't be a bubble without this sense of rationality.

(Edit: I wasn't clear. This is the same reason it's dangerous to short bubbles. You have no idea how high the price is going to go - a bit higher, a bit more, 2x, 4x, 8x, 16x, etc.)

Less risky plays include:

  • Buying out-of-the-money calls to play the bubble long. You can still exploit the rise in the bubble if it continues. And your downside is limited if it doesn't. The price for this protection is the time premium on the option. (Edit: Actually, the price for the downside protection is the price of the option, not just the time premium).
  • Buying out-of-the-money puts to bet on the bubble imploding. Loss is limited if the bubble continues to rise. If the bubble bursts, you make money.
  • Do both, especially in a period of relatively lower volatility. You're playing the return of or increase in volatility here.

As I said above, I have no idea if oil is in a bubble (though the prices are sure acting like it), so the safest strategy is to do nothing. OTOH, if I had to play the NASDAQ bubble again, I'd buy out-of-the-money calls. I certainly would not try to time a bubble-burst.

BTW, the greatest (luckiest?) trade I have ever seen was someone buying far, far out-of-the-money calls on CMGI in late 1999. The stock exploded and tripled in less than a month. He turned $1000 into $1 million in less than a month. It was part luck (he picked the right stock) but also part strategy (playing a bubble correctly with minimal downside risk). I saw this in realtime so I know he didn't fake it.

So if I understand you

So if I understand you correctly, you're saying that the problem with putting your money where your mouth is in calling a price rise a bubble is that we don't know when the bubble will burst? But then why not just sell short over a longer term? Is it because over the long term, we expect the prices of oil (and real estate) to rise anyway?

I guess that makes some sense, but I still have a problem with people claiming to know more than what the market knows but not being able to profit from their knowledge claims. If you're unwilling to speculate with your own money, it seems like you shouldn't be claiming that you know more than the speculators.

Selling Short

You can't sell short over a specified term. What you do is borrow stock at the market interest rate, and then sell it. You have to buy it back at some later point, and pay the interest as you go. If at any time the market price of the stock rises to the point where your liability is equal to the amount of cash in your brokerage account, your broker will force you to buy the stock, wiping out all your cash. It doesn't matter if you're proven right in the long run. As Keynes said, the market can stay irrational longer than you can stay solvent.

As Jonathan points out, buying put options can limit your risk, but long-term options are expensive, and short-term options are likely to expire worthless unless you time it just right.

Example

Let's say it's 1998, and you're convinced there's a bubble in tech stocks. You see this company "AOL" that caters to unsophisticated users, sends them free CDs in the mail to trick them into signing up for modem access, and widespread broadband access via cable/DSL is right around the corner. You believe that at best the stock deserves a price of $25 but the current price is $50.

Should you short AOL?

Suppose you do short AOL. A week later, the price is $55. Now, you're down $5 and you're a bit nervous. A week later, the price is $60. A month later, the price is $75. The butterflies are really flapping now. The very next day, the price jumps to $90. Now, you're down $40. Are you going to have the courage to hold?

The price rises to $125, then to $150. Are you still going to hold?

Because of the bubble characteristics, the price rises to $1000 before the bubble eventually bursts and settles to the "fair" price of $25. Obstacles to making a profit on the trade include:

  • You have to be able to hold through a parabolic rise with conviction in yourself. Most people can't do this.
  • As Brandon said, as the price rises, your broker will probably buy back the shares (like a margin-call in reverse), even if you have conviction.
  • Earnings might 'catch up' with price. The stock might be fairly priced at $25 in 1998. But by the time the bubble bursts in 2000, earnings have increased to the point where fair price is now $75. Even after the bubble pops (and you have no idea when it will), you'll still be down $25.

This AOL example is a real-life I example I witnessed personally.

I guess that makes some sense, but I still have a problem with people claiming to know more than what the market knows but not being able to profit from their knowledge claims. If you're unwilling to speculate with your own money, it seems like you shouldn't be claiming that you know more than the speculators.

This is a very good argument, and it gets to some fundamental issues that IMO are unresolved.

1) I'm not sure knowledge/information is the only determinant of price. Over the long run, I believe it is, but in the long run, we're all dead. By the time that information:price fidelity (basically, efficiency) is restored, you could lose your shirt.

Emotion, social proof, and herd behavior are also important determinants of price in the short term. (And this is the only reason I play the market: a belief that I can control my emotions better than the vast, vast majority of market participants. If I didn't believe this, I would not play.)

2) The common way to think of efficiency is "the price is very information rich". Another way I think of efficiency is "it's very, very, very hard to beat the market". IOW, it's always hard to beat the market, bubble or not. When the bubble exists, lots of people know it's a bubble, yet almost nobody makes money, and the ones that do are mostly lucky. The information is already out there, yet the bubble keeps growing. I can't really conceptualize a good way to think of this. It seems like infinitely nested loops of "I know that you know that I know..."

3) The question is raised - "Should anyone ever play the market, bubble or not?" I'm not sure of the answer, but I lean more and more in the direction of "no".

Taleb's methods?

I think Taleb (the black swan guy) has a way of betting so that he loses small even if prices go sky high and wins big if prices drop through the floor. (This is possible because the small losses are almost certain and the huge wins are unlikely - but maybe his method doesn't work for bubble bursts because bubble bursts aren't black swans...?)

Yeah

That's basically very similar to one of the methods I described in my original reply to Micha: buy a small number of way out-of-the-money puts each month knowing that you'll most likely lose money on them. But if the bubble bursts, you make a killing. Of course, Taleb's methods would be much more complex, as he has more sophisticated modeling and derivatives at his disposal.

Like I said earlier, I'd do the opposite of this method: buy way out-of-the-money calls on market 'corrections' betting that the bubble will resume. Repeat as long as bubble keeps growing. When the bubble actually bursts, the 'correction' will keep imploding and I'd lose on that trade, but hopefully, I'd make money many times before that happens. And the loss is minimal anyway because out-of-the-money calls aren't a lot of money.

But it wouldn't work to correct the market

So you're saying that even these safe bets wouldn't work to correct the market - that is to say, the bets on the bubble bursting, which presumably would tend to push the market down, wouldn't tend to outweigh (in their effect on the market) the bets on the bubble continuing/resuming after a correction, which would presumably tend to push the market up.

I'm not sure I understand your question

Are you asking what influence puts have on the price of the stock?

If so, I don't know the answer to that. When someone sells or sells short a stock, they're exerting influence on a fixed quantity of stock issued by the company itself. So there is an effect.

When someone buys a put, they're buying a contract to be able to sell a stock at a given price in the future. But the contract is created on the other end by someone who's willing to sell such a contract. He's betting the opposite as you.

Intuitively, I think there would be some sort of influence, but I don't know what it would be. I think the overall effect would be a change in price such that as many options contracts as possible expire worthless.

My heuristic

My heuristic is that if you bet that the market will move in a certain direction, you are going to move it in that direction in the short run. This seems to be what normally happens. Think of this as a bit like the no-perpetual-machine principle. Purported perpetual motion machines can be difficult to analyze, but even without analyzing them we can pretty safely predict that they will not work. Now, I'm sure that the no-perpetual-motion-machine is more than just a heuristic, I'm sure there's a way to derive this rigorously from laws. So it's just a loose analogy.

[ADDED TEXT: So no, I am not asking what direction the effect will be. I am assuming it will tend to be in the direction of the bet. Since you don't feel certain about that, then you can't really answer the question I was asking, which was whether these safe bets might help to burst the bubble/prevent the bubble from getting bigger sooner rather than later.

The evidence that bubbles exist suggests that they would not help, but up until today I had a completely different interpretation of bubbles. Not that people knew they were bubbles but were unable to burst them early on. But, rather, that there is a deep epistemic problem in identifying bubbles, which can only reliably be recognized as bubbles in retrospect.]

But the contract is created on the other end by someone who's willing to sell such a contract. He's betting the opposite as you.

Yes he is, but he was there before you got there/would have been there whether you entered or not, and would have traded with someone else. By entering the market you affect the market in a certain direction. Consider an analogy with regular old banana markets. If you enter as a buyer who values bananas enough to buy at the market price, you (ever so slightly) drive up price. This is true even though when you buy, you are trading with a seller who is affecting the market in the opposite direction. That seller is there independently of you, so even if his effect counterbalances your effect, still, your entering the market has an effect in a certain direction as compared to an alternative world where you do not enter the market but the seller is still there trading with someone else.

Puts and calls are

Puts and calls are replicated using dynamic hedging. When you sell a put, 99.99% of the time, the other party will be a bank.The first thing they'll do once the put is sold is to short the underlying contract to hedge themselves. They will change the amount by which they are short every day, in order to neutralize the risk. The other 0.01% of the time, you change the price of puts anyway, and if they go to far from the price of replication, there's an arbitraged triggered.

In practice, if you sell a good amount of puts, you will actually often see the sell orders on the underlying contract being placed on the market.

Ah

Well, I can't say I learned anything from this, because your comment presupposes that the reader has passed finance kindergarten which I have not attended yet, but I was suitably impressed.

Yeah, Arthur's knowledge of finance is impressive

I'm betting (ahem) that he works in the finance industry.

I'll bet against you for a

I'll bet against you for a million dollar (and leave my job :p )

I'll explain delta hedging in a few words :

You have an asset with a price S(t). A european call (it's a type of call option) with maturity T and strike K pays you at time T max(S(T)-K,0). So let's say that the strike is 90 and the maturity one year. If in one year the underlying (the asset, S) is worth 100, then the owner of the call receives 100-90 = 10. If the asset is worth S = 80, the owner receives max(-10,0) = 0... nothing. So you get whatever is above the strike.

What should the price of a call be ? In order to find out we work backward. If we were 0 days to maturity, the value would be known... it's the payoff max(S(T)-K,0)... easy. Now assume the call you sold to a client will expire tomorrow... today, the stock happens to be at 90. You think tomorrow it will either be at 100 or at 80. You don't need to have a view on the direction the price will take, only on the expected magnitude, the volatility. Anyway, if the price goes to 100 you'll have to pay the client 10, if it goes to 80, you'll have to pay 0. What do you do? You buy 1/2 asset ( of course you are covering thousands of options, so you simply take that number divided by two ). If it goes up by 10, you'll make 5, and will have to pay 10, if it goes down by 10, you'll lose 5. In every cases, tomorrow you will need to pay 5, regardless of the direction the stock takes : you are hedged... working backwards one can find the price of an option using this covering strategy. The generic requirement is that you should allways own
delta = dC/dS units of the underlying, where C is the price of the call, and S the price of the asset... whatever you make or lose on the call because of a movement is S should be compensated by your position in S.

The big assumption is that we know the magnitude of the moves... an option price strongly depends on the expected volatility of the asset.

To put it in a nutshell, in theory, selling options is not the same as selling insurance. When you sell insurance, you offer to take some risk, at a premium. When a bank sells an option, it is selling a strategy. The price of the option is linked to the price of the underlying because one can use the delta-hedging.

When you buy puts on oil, the bank starts selling oil future to avoid taking any risk.

I admit that I don't have a

I admit that I don't have a very good grasp on investment tools. But even given your three bullet points, it seems that someone wealthy enough and confident enough that a high price is due to a bubble should be able to sustain holding short for a long enough period of time to make a profit.

Same here

Jonathan's thesis seems to be that bubbles are recognizable early on but that certain properties of bubbles prevent investors from capitalizing on their knowledge (and, as a side effect, deflating bubbles). An alternative view of bubbles which I have seen before and which I had (maybe wrongly) thought was the commonly accepted view, is that bubbles can't be recognized as bubbles until after they burst.

Now, some evidence against this position is, "during a bubble a lot of people loudly proclaim that it is a bubble."

Doubtless. But this only means that they believe that it is a bubble, and (if it is) that they happen to be right. This doesn't prove that bubbles are genuinely recognizable and knowable except in retrospect.

Jonathan's thesis is an alternative to this. Jonathan pointed out that it can be extremely dangerous to try to capitalize on your accurate knowledge that a bubble is a bubble. You could be ruined if the bubble goes up too far before it bursts. He writes:

The very next day, the price jumps to $90. Now, you're down $40. Are you going to have the courage to hold?

If it takes courage to try to profit from your accurate knowledge that a bubble is a bubble, then it must be dangerous to do so. In answer to this suggestion, I offer up the possibility of safe bets, a la Taleb, which do not require courage but which might - if enough people make those bets - have the same effect of bursting the bubble early. The discussion however, seems to have stalled, because I don't seem to have expressed myself very well. I thought that Jonathan was hinting that the corrective effect of Taleb-like bets would be cancelled out by the anti-corrective effect of reverse Taleb-like bets (which Jonathan claims that he himself would choose). He writes:

Like I said earlier, I'd do the opposite of this method

I attempted to ask Jonathan point blank whether he was saying that my attempt to get around the "courage to hold" problem via safe Taleb-like bets doesn't work, because reverse Taleb-like bets are even more tempting. But I don't seem to have communicated my question.

Ah, I didn't realize that's what you were saying

In answer to this suggestion, I offer up the possibility of safe bets, a la Taleb, which do not require courage but which might - if enough people make those bets - have the same effect of bursting the bubble early. The discussion however, seems to have stalled, because I don't seem to have expressed myself very well. I thought that Jonathan was hinting that the corrective effect of Taleb-like bets would be cancelled out by the anti-corrective effect of reverse Taleb-like bets (which Jonathan claims that he himself would choose).

I wasn't hinting at that. I doubt anyone has the power, even in large numbers, to push the market in a certain direction. In fact, one of my fundamental beliefs is that only contrarians make money. My positions are always against the herd. The greater the degree to which they are against the herd, the more likely I am to make money. So I would not take a position believing that I could push the market in the direction I am betting, but rather than the number of people who are betting in the opposite direction is soon to be exhausted.

The strategies I described (both of them, Taleb-like and anti-Taleb-like) are purely about playing probabilities with minimal risk, not about exerting influence on the market, merely about going along for the ride.

On another note, I've found this discussion very stimulating. Kudos to Micha and Constant. Talking about bubbles and making money in the financial markets is one of my favorite topics. I was similarly stimulated (don't go there Scott) by the Peter Thiel piece Patri linked to a while back.

I think the next bubble we'll see is a gold bubble (as part of a generalized commodities bubble). It's a decade in the making so far, and probably will take at least another 5 years to reach frothiness. I can't wait to play it. If this blog is still around, I'll post my trades in real time.

I don't think the immediate

I don't think the immediate goal for any individual investor is (or should be) to shift market prices. That seems like it would be a secondary outcome of investment decisions. But if the large number of people who claim to be certain it's a bubble were able to put their money where their mouths are, this seems like it would have a bubble-deflating effect. Sort of a reverse bubble, to balance against the straightforward bubble.

In some blog comment thread I read recently, someone said the Saudis have been investing in gold lately.

For what I consider above-average knowledge in economics, my understanding of finance and investment tools is piss poor. Any good books or articles you guys would recommend?

I mean unintentional, side-effect push

I didn't mean that people would actively try to end bubbles by purposely pushing them. I was appealing to the idea that people inadvertently push stock prices through their actions.

I doubt anyone has the power, even in large numbers, to push the market in a certain direction.

The market is nothing but people, so people do (inadvertently) push it in all the directions that it goes. There is nothing else but people to push it. I think you're talking here about the difficulty/impossibility of a group of people to deliberately shift the market, but that's not what I was talking about.

So I would not take a position believing that I could push the market in the direction I am betting, but rather than the number of people who are betting in the opposite direction is soon to be exhausted.

I didn't mean to suggest that you would take a position in order to move the market, but merely take a position in order to make a profit. You, along with everyone else in the market, inadvertently by all your actions determine where the market goes. The heuristic, rule of thumb that I mentioned was that each action in the market contributes a tiny amount to the way the market is pushed, and that a bet that the market will go up, however that bet is realized (be it a straight stock buy or something complex) will tend (ever so slightly) to push the market up. That is, if you think that a stock will go up, and you buy it, this will increase the demand for that stock ever so slightly by adding your own demand to the whole, which will drive the price up ever so slightly (the change may be merely probabilistic - it may take many of you to actually drive the price up). My simple rule of thumb (because it seems to capture many different kinds of behavior) is that however you realize your bet, be it a purchase or a sale or an option or whatever, you will tend to push the market in the direction you are predicting it will go in, not because you want to push it, but inadvertently.

The strategies I described (both of them, Taleb-like and anti-Taleb-like) are purely about playing probabilities with minimal risk, not about exerting influence on the market, merely about going along for the ride.

I know - I didn't mean that the intention of such bets was to correct the market, but merely that the cumulative effect of all the activities is in fact what drives the market, since the market is nothing other than the market activities. So, regardless of your intention, every action affects (or has some probability of affecting) the market.

Ah

Effect not intention.

Still, I don't know the answer to whether numerous Taleb-like bets would influence the market. But why is the question important? Would it change your strategy?

The question seems important

The question seems important to me, and underlies my original purpose in starting this thread. If there is nothing skeptics can do, even unintentionally, to stop bubbles when they think they exist, that seems unfortunate, because not only are the pro-bubble investors going to eventually lose out, but regular consumers of the commodity in question will lose out too for the duration of the bubble. I don't mind fools losing their money for making silly investment decisions, but it's a shame if this has negative side effects for lots of non-fools. And it also deprives us of valuable information about the future, since outside observers will have a hard time knowing if prices reflect genuine information or herd beliefs.

Ditto more or less

Ditto, except for the part about giving a damn. :-)

Good points

The question seems important to me, and underlies my original purpose in starting this thread. If there is nothing skeptics can do, even unintentionally, to stop bubbles when they think they exist, that seems unfortunate, because not only are the pro-bubble investors going to eventually lose out, but regular consumers of the commodity in question will lose out too for the duration of the bubble.

Wasn't your intention to ask why skeptics don't put their money where their mouths are? This question centers on personal gain. Yet the quote above centers on societal utility.

There are lots of theories about why bubbles form, and one of them is the Austrian theory that bubbles are a result of an unnecessary increase in money supply. The money has to go somewhere, and it 'falsely' elevates prices of whatever sector the bubble is in.

Another theory that I've put forth before is that bubbles form due to moral hazard.

Skeptics can help prevent bubbles from forming in the first place by calling attention to increases in money supply and moral hazard.

Another theory is that, every so often, people just go batshit insane. Not much can be done about this, but at least we can document the insanity for future generations.

It could be none of the above, in which case, for me at least, societal utility will take a back seat to personal gain. I'll either stay out of the market altogether or use one of the small-bets-for-large-gains strategies.

This question centers on

This question centers on personal gain. Yet the quote above centers on societal utility.

Well, yes. That's sort of the point of markets and market prices. They are determined by the self-interested actions of many separate individuals, but as if "led by an invisible hand to promote an end which was no part of his intention."

Well, yes. That's sort of

Well, yes. That's sort of the point of markets and market prices. They are determined by the self-interested actions of many separate individuals, but as if "led by an invisible hand to promote an end which was no part of his intention."

I realize that, but I see them as separate questions. A common theme in prediction markets is this separation. Topics can be divided into, "Why would the average person play market X?" and "What is the social utility of market X?"

Example:

"Why would the average person ever play a weather prediction market?"

Answers: He wants to make money. He thinks he can use the NOAA data better than the NWS. Etc.

"What is the social utility of a weather prediction market?"

Answers: Weather forecasting better than the NWS, conventional meteorology, etc.

Similarly, the questions, "Why don't skeptics short the bubble?" and "What is the best strategy for a skeptic to use?" are different from "What can the skeptic do to help soften the damage done by bubbles?" and have different answers.

I guess I still don't

I guess I still don't understand what you are getting at. The answer to "What can the skeptic do to help soften the damage done by bubbles?" is: pretty much nothing, if the market is large enough and skeptic isn't ridiculously rich already. I am interested in the question, "Why don't skeptics short the bubble?", because it seems like if they could do so successfully, they could get rich while also (if there are enough of them) softening the damage done by bubbles.

The Power O' Skepticism

"The answer to "What can the skeptic do to help soften the damage done by bubbles?" is: pretty much nothing, if the market is large enough and skeptic isn't ridiculously rich already."

Didn't you already learn this lesson watching people go to church on the off chance they'll win an eternity of bliss? The way for the skeptic to make money off that is to become a televanglist.

Curiosity about bubbles

The underlying question is, "what is the nature of bubbles". I'm not investing, I'm just trying to understand them. What a bubble is depends on how it works - e.g., why it appears, why it takes a while to burst, what sorts of trades have what effect, etc.

While I certainly don't

While I certainly don't believe this is currently the case, if someone cornered the oil market (big if), you couldn't really arbitrage it like that, your short sell would be squeezed out. The only people who could resist it would be the oil producers by refusing to sell and waiting for the guy to go broke on interest and storage.

A note on terminology

The use of "Taleb-like" and "anti-Taleb-like" are mere attempts at analogy: small bets that frequently fail but infrequently have large payoffs and vice versa, respectively.

They probably don't have much to do with actual Taleb strategy. If Taleb were to show up on this thread, he'd probably say, "WTF are you guys talking about?!? This has nothing to do with my strategy. I'm going to do a thousand pushups because I haven't done any in the last few months! Don't you dare think about touching my salad!"

The word you're looking for

The word you're looking for is skewed.

Taleb is not about so much about skewed distribution, rather about fat tail distribution. However, betting that a distribution has fat tail is itself a skewed strategy (it loses frequently and gains a lot rarely).

Since most investments are made by people with assymetrical incentives (traders, funds, mutual fund, etc) I believe that bets with negatively skewed return are probably overbought and bets with positively skewed return oversold. In that case, an individual, with linear incentives could make a profit by investing in right skewed bets, for example betting that distributions will display fat tails.

This is exactly what Taleb advocates, although I believe he does it mostly for epistemological reasons while I think the systemic explanation could be more adequate.

Imagine a poker table where the players where playing for clients, with a guaranteed share of the profits and no downside... they'd play all play very lose, and you could make money by being tight... although their style would be correct, based on their incentive.

One of the most skewed bet right now is the Yen, Australian dollar carry trade.It makes a fair amount of money every month, something many people are happy to show to their boss or client. Conversely, few traders would go against that trade because they would stand to lose money for a very long time and could be fired etc. In this case, betting on a sharp appreciation of the yen relative to the dollar probably has a positive expectation.*

* Or I could just be mumbling crap, invest at your own risks, etc.