Redirecting

Thursday, September 02, 2010

What I'm Reading This Week


"Does it matter who controls the businesses of the country?  Does it matter who regulates the businesses of the economy?  Should these people be smart or dumb?

One cost of the meddling that the Fed and Treasury have done through the bailouts is that dumb people are left in place.  People who mismanaged their firms still manage them, and regulators that misregulated are still in their jobs.  Both are rescued by taxpayer largess, but it reveals another hidden cost of bailouts — they make us less competitive/effective as a nation, because we do not let  firms fail, and we don’t fire regulators that were negligent, including the Fed.

The optimal outcome would be for bright managers to buy failed firms out of bankruptcy, and for failed regulators to be replaced with new ones that have a chance of doing things differently.  Aside from that, if you never fire regulators for failure, you will never motivate them to do what is right.

And this is true of most meddling by the Fed or the Treasury, picking favorites, not letting bad firms or regulators fail.  This extends to QE.  If you need lower rates in order to survive, you probably don’t deserve to.  All a lower rate structure does, if the government or Fed is forcing it, is encourage investment in lower yielding investments, because they can be financed cheaply for now.  This is an aspect of the liquidity trap, and the Fed is deepening it with their policies."


Extend and pretend anyone?  We've talked about this concept ample times already here: Greece solved nothing - they just delayed the day of reckoning for their problems.


"There are two issues here: The first is “why did Lehman collapse?” The second is “Why didn’t the Fed rescue them?”"

- Jeff Matthews with more on "Executing Strategies"

-Nemo with an interesting post on expectations; specifically, inflation swaps.  He even indulged my curiosity in the comments (still ongoing as I publish this piece)

- The legendary Dr. Pauly guest-blogging for AlCantHang at the FullTiltPoker blog:


"What the hell is a “durrrr” anyway?

My gut told me the word was made up, but a part of me thought that maybe durrrr had origins in dead languages like Latin or Mayan, or perhaps it was the scientific name for a breed of howler monkey. After a thorough internet search, I discovered… nothing. Indeed, durrrr is a word wholly concocted by Tom Dwan. In an interview, he once mentioned that he made up the “durrrr” moniker in an attempt to annoy the hell out of his opponents. He was spot on, because nothing can tilt you more than knowing that you’re losing a few buy-ins to a guy named durrrr, a word that when said aloud resembles the sounds that an incontinent person makes before they defecate themselves."


-KD

20 comments:

getyourselfconnected said...

Cool stuff at nemo's site but a tad over my head!

I was on board 100% with your Apple post before.

Blurtman said...

One common misassumption is that leaders are smart and competent. Nope. What they are good at is getting into positions of power.

Hemant said...

On the Nemo running debate - think you are almost right, and Nemo is definitely wrong.
first, look at nemo's coin flipping argument.
(i)This is certainly wrong. If someone comes to me and asks me to make a price, I will show 0.45 / 0.55. Which means I will have the edge as market maker. if I go to someone and ask him to make a price, I will pay the bid/offer.
(ii) second, Nemo seems to argue that one party in the coin flipping contest is getting certain cash flows, and the other uncertain cash flows. This is wrong - both parties are exposed to uncertainty - one receives 1- if(head,1,0), other the inverse of this.


On general financial instruments, lot of research on the topic of forwards vs. expected spot. Keynes gave your hedging argument in 1930s to justify that forwards can be higher or lower than expected spot, depending on preponderance of hedgers and market makers. Modern theory has a more confusing explanation - in terms of "market price of risk", a completely unobservable variable.

Kid Dynamite said...

Hemant -

I think Nemo might respond to your first part by saying "you will make the market 45-55 as a market maker, but you won't find people willing to pay 55, you'll only find people willing to sell at 45"

I like the point in your second part of your comment.

I also think that I explained in one of my last comments last night on his site why the coin flip scenario and the "fire insurance" analogy are misplaced - it has to do with boundaries of payouts.

Greycap said...

Hey KD, I tried to post at self-evident but couldn't.

You are making an argument based on supply and demand: if more people want to pay inflation than receive it, then that should push down the price of inflation, right? Yes. I suspect that Nemo would agree with this.

But the structural interest in the market is not related to what people expect to happen. There are natural inflation payers, with nominal liabilities but real assets (e.g. the government), and there are natural inflation buyers, with real liabilities but nominal assets (e.g. pensions.) It happens that the latter outnumber the former, and this is not going to change just because we get deflation.

What you are talking about is different: you expect deflation, and you want to place a bet on this expectation. But your bet will only pay off if there turns out to be more deflation than contracted for in the bet. If the market implies more deflation than you expect, you just won't place your bet - you'll just buy nominal bonds at 0%. The only reason for entering a money-losing bet is to eliminate risk. That's it. This is Nemo's point.

Greycap said...

Some other remarks. It depends what you mean by "systematically" - maybe your definition is "always", 100% of the time. But that is not how most people use the term - to the rest of us, it means "usually", that there is a bias.

Under the normal definition, swap rates do in fact systematically overstate forward rates. All interest rates do - that is because long rates are riskier than short rates, and the price of risk is positive. This is true even when the curve is inverted! An inverted curve is the IR analogy to the deflation case you are thinking of - you expect rates to fall, but you wouldn't buy the long bond at a price that implies rates will fall by more than you expect.

I don't like Nemo's "fixed payment" way of explaining the issue, because it really is just about risk. For instance, a CDS protection buyer pays a fixed coupon in return for an uncertain payment associated with a default event. But it is the protection seller that is taking the risk, and default probabilities backed out of CDS spreads (or bond spreads, for that matter) are systematic overestimates.

Kid Dynamite said...

Greycap -

you wrote "It happens that the latter outnumber the former, and this is not going to change just because we get deflation."

but it doesn't always have to be like that, right? in fact, isn't is possible that the former could outnumber the latter, reversing the direction of the premium? Why is it crazy to think that there could be a change in the balance of inflation payers vs buyers, tipping the equilibrium in the direction of the payers, so that the contract will understate the level of inflation instead of overstate it? One of the commenters on Nemo's site linked to a paper saying this is exactly what happened in the early 2000s. (again, this is where I feel like I may be missing something specific to inflation swaps - i'd love it if you brought me back to an equity example)

you also wrote "If the market implies more deflation than you expect, you just won't place your bet - you'll just buy nominal bonds at 0%." What if nominal bonds aren't available at 0%? I don't want to get bogged down on that point, though, because I want to ask you this:

you wrote "But it is the protection seller that is taking the risk, and default probabilities backed out of CDS spreads (or bond spreads, for that matter) are systematic overestimates. "

Isn't Nassim Taleb's entire thesis of existence that we systematically underestimate tail probabilities? Is that related somehow?

can you draw an analogy for me to the equity markets?

Kid Dynamite said...

Greycap, yes - I agree with you on "systematically" - it doesn't have to mean 100% of the time - and I'll even agree that inflation swaps usually over estimate the rate of inflation, for the reasons you and Nemo explained. But that doesn't mean it's always like that, right? The balance can shift (As in my previous comment here) and reverse the risk premium, right? I think that's my only point. It sounded to me like Nemo and the Prof were saying "it's always like this, because..." and I was just thinking "no, that can't be true. it can't ALWAYS be like that."

Kid Dynamite said...

Greycap:


I have another potentially tangential/irrelevant question:

isn't it possible that before the shit hit the fan in 2007/2008, that CDS prices systematically UNDER estimated default probabilities? Since the balance was weighted heavily toward the AIGs of the world who were eager to sell these "riskless" contracts and get paid to do so? (remember Cassano's quote about not being able to construct a single scenario where they would lose even one dollar on their trades???)

Greycap said...

1. From a nominal dollar perspective, real dollars are risky. From a real dollar perspective, nominal dollars are risky. There's nothing crazy at all in principle about thinking there could be more people with the first perspective than the second, but Nemo is right to say that it is natural for people to care more about what a dollar buys than the dollar itself.

Either way, it doesn't matter whether the risk is on the upside or the downside - you always pay to get rid of risk. So inflation vs deflation makes no difference.

Re: 0% bond, the higher the interest, the better for you, as always. I'm just saying that you have no reason to accept a negative rate that exceeds your expectation of deflation unless you have a nominal dollar perspective and are paying to get rid of risk.

Re: Taleb, I don't want to go there in a blog comment - life is too short.

Re: equity markets, you have to distinguish between transient liquidity effects and multi-day investment horizons. You can't say anything meaningful about expectations just because liquidity dries up and a 1c trade prints, like in your HFT examples. But would you buy an equity for more than you expect to sell if for? Would you sell short for less than you expect to buy it back for? No, not unless by doing so you were locking in a profit on your total portfolio - i.e. hedging risk.

Greycap said...

KD, I completely agree that the market can make mistakes - I'm not one of those strong EMH nuts. So yes, absolutely, credit spreads can be too low, and obviously were on MBS/ABS securities. Likewise, implied forward rates can be lower than rates eventually turn out.

But the theory Nemo is referring to is that credit spreads imply higher default probabilities than what people think are the real probabilities, and the IR curve implies higher forward rates than what people think rates are going to be. Remember, nobody knows for sure the real probabilities of any of these things! They are willing to deal only when they think it is profitable - except when they are shedding risk. At least, when they think they are shedding risk - they can be wrong about that too!

Kid Dynamite said...

Greycap wrote "But would you buy an equity for more than you expect to sell if for? Would you sell short for less than you expect to buy it back for? "

right - of course not - so who is the "shedder of risk" in this scenario? the buyer of stock? or the person who shorts it to him???

neither? so then the theory doesn't apply to equities? I guess I look at an equity trade (buyer vs short seller) as just another "swap" - of payments...

Kid Dynamite said...

Greycap: I think I just had an epiphany:

Nemo is saying that the implied prices in inflation swaps overstate the MARKET's true expectations of inflation, while I'm saying that the Market's expectations of inflation may over or underestimate the actual eventual levels of inflation ???

Greycap said...

KD: right - of course not - so who is the "shedder of risk" in this scenario? the buyer of stock? or the person who shorts it to him???

They are both taking risk! And they are both being paid to do so according to their own expectations.

But the premium on the security depends on the aggregate expectations of all participants. And an equity market cannot function without net long interest - if people were not invested net, there would be nothing to short. So equities always trade at a discount to what the market as a whole thinks they are worth. The fact that markets are often wrong makes no difference.

Kid Dynamite said...

Greycap:

"They are both taking risk! And they are both being paid to do so according to their own expectations."

Gahhh! - but that's pretty much my point. EVERY trade is like that, not just equities... In every trade two parties are taking risks. The fact that in general, one of those sides usually consists of people who are trying to reduce risk doesn't mean that this is ALWAYS the case, right?

Do you think the paper that JaguarRacer linked to in Nemo's post was wrong when it said that sometimes the inflation swaps go the other way and underestimate the true levels?

I actually thought that equities trade at a premium to what people think they are worth - because people are stupid, they don't even "think" at all - they just invest by mandate. Forced investment means that they are willing to pay more than what they think they are worth, but that's a sublime concept because they don't think at all! (I'm thinking largely of retirement/401k accounts here, etc)

Kid Dynamite said...

actually, Greycap, I want to discuss that claim some more: "So equities always trade at a discount to what the market as a whole thinks they are worth."

do you really mean ALWAYS? or do you mean the vast majority of the time... I'm thinking of Sep 2008, March 2009 in particular as exceptions...

and I think it's very much related to my thoughts on inflation swaps.

Anonymous said...

Just to add to that inflation swap debate. I think Nemo's arguments make no sense. As you say, one trade in isolation tells you nothing about which side is shedding risk or assuming risk -- both may in fact be assuming or shedding risk. If I am long nominal dollars, I may be willing to pay away some money to convert them to fixed dollars (i.e. accept a payout that overestimates inflation), conversely if I am long real dollars and want nominal, I may be willing to pay away some money to convert (i.e. accept a payout that underestimates inflation). The determination of which predominates in the observed price of inflation depends on market structure and the population of participants.
His comments about equity options make even less sense. He seems to be saying that because options are always positive premium, risk premiums are always positive. But the premium of an option is largely not risk premium, it is expected payout. And again, either the buyer or seller or both may be shedding or assuming risk. If I am already long the option, of course I might be willing to sell it at less than what I expect the payout to be, just to make that payout certain. Or if I am short, I would be willing to pay more than fair value to cover.
Or to take the example of the coin-flip, it surely makes no sense to argue that if it trades at 35c, that is a 35c risk premium -- no, that is a negative risk premium of 15c. The only reason we find it difficult to believe that a coin flip can trade at anything other than 50c is because this is an artificial situation where the probabilities of heads and tails are known with certainty.

Kid Dynamite said...

so yes, Anon - the fact that in the coinflip example we know the probabilities with certainty bastardizes it to a state where it gets confusing to talk about real applications. In the real world, it's really hard to determine where the pricing factors come from: is it differences in expected value? or differences in risk.

your comment "The determination of which predominates in the observed price of inflation depends on market structure and the population of participants" is precisely my point. I think that Greycap and Nemo are saying that IN GENERAL, most of the time, it leans one way - which I agree with - but I'm saying that it can lean the other way too. USUALLY people are risk reducing by taking a certain side of the inflation swap (inflation buyers), but I don't think that is an absolute factor - it doesn't have to be the case.

I wanted to ask Greycap yet another related question on that subject: if Obama sent everyone in the country an envelope that contained either $1 or $0, with equal probability, we know that the EV is 50c. I would be happy to act as market maker, paying 45c or selling them for 55c.

I think that Greycap is saying that plenty of people would be willing to sell their envelopes to me at 45c - accepting less than the fixed EV, because they are reducing risk (the risk that they might get NOTHING). THat's true - but they are also eliminating the possibility that they get the big fat $1 payout. Utility curves come into it. I think that Greycap is also saying that although I'd be happy to sell my envelope for 55c, that no one would be willing to pay me 55c for it.

I disagree with that assertion, and it may be where classroom economics diverges from the real world. Nemo touched on it briefly in his thread when he mentioned the possibility of people becoming risk SEEKING when faced with the possibility of certain loss - it's related to that. I don't think it's hard to imagine "irrational" people paying 55c for my envelope. Why? maybe they are desperate, and the 55c they have will do them no good, but a dollar would help them... that's probably a twisted way of saying that utility curves are not symmetrical.

there's an even easier explanation for my claim though, providing evidence that it's correct (that people WOULD pay 55c for something worth 50c) - that is the existence of lotteries (and casinos too)... They exist, so by definition, people are willing to pay more than EV, even when they aren't reducing risk. I'm wondering what Greycap would say about the existence of lotteries, and how they apply to the claim "The only reason for entering a money-losing bet is to eliminate risk. "

Anonymous said...

Kid, agree completely. And I just realized Nemo's got a sign error in his discussion of your example of a home-owner hedging deflation risk. The homeowner wants to receive a fixed nominal amount (presumably because that's what he needs to pay his mortgage), and will be willing to accept an implied inflation rate of -31% to protect himself, exchanging expected -30% for certain -31%.

Greycap said...

do you really mean ALWAYS? or do you mean the vast majority of the time... I'm thinking of Sep 2008, March 2009 in particular as exceptions...

No, not always, not in a panic, when nobody knows what they think other people think the market is worth. But then can it mean anything to talk about a "market price?" Also, for the price of risk to have meaning, there has to be an alternative to measure the risky asset against, some reasonable traded proxy to a risk-free asset, such as an overnight account.

I actually thought that equities trade at a premium to what people think they are worth - because people [...] just invest by mandate.

Sorry, but unless you think that mandate is about to end, that's self-contradictory. It's exactly Ritholtz's example about "perma-bulls wanting to short." The price of an asset is still what other people will pay for it.

Look, in a world where people cared more about the number in their bank account than what that money can buy, then real dollars would seem risky and nominal dollars safe. In that case, inflation swap rates would understate market expectations of inflation. The same thing would be true if people cared more about how many units of S&P 500 they owned than about the price those units would sell for - equities would be safe and money would be risky, so equity prices would overstate their value in money terms.

In real life, in America in 2010, you should assume the opposite.