money love

by russell

OK, this is a short one.  I'm trying to get my head around some numbers and it's making my brain hurt.

This wiki article cites this Bank for International Settlements report to say:

Thus, derivatives trading – mostly futures contracts on interest rates, foreign currencies, Treasury bonds, etc. had reached a level of $1,200 trillion, $1.2 quadrillion, a year. By comparison, U.S. GDP in 2006 was $12.456 trillion.

For "derivative" I read "a financial instrument based on speculation about what the value of other things will be".  AKA, a bet.  A well-informed bet, perhaps, but a bet.

Please correct me if I'm mischaracterizing the nature of derivative financial products.

So, in 2006, the market in speculative financial instruments was approximately one hundred times the value of the entire GDP of the United States.

One hundred times the value of the entire GDP of the United States.

US GDP runs in the neighborhood of 30% of global GDP, so that means in 2006 the market in derivative products was, order of magnitude, thirty-plus times the productive output of the entire world.

Wait, what?!?

It's late in the day, I'm a little tired, maybe my math is wrong.  Somebody help me out if so.

If it's not, does anyone besides me see a problem here?

 

Comments

[Nate]

No, that's pretty much right, they made bets for more money than actually existed in the world.

And raked in a nice commission from each of them.

[cleek]

the title "Master Of The Universe" does have its perks

[Peter Leppik]

This is very misleading, even if technically correct.

In most derivative contracts, the actual amount of money which will ever change hands is only a minuscule fraction of the "notional" value of the contract.

In an interest rate swap, for example, one party will pay the other party the difference between two interest rates times the face value of the contract. Since the two interest rates are normally expected to be very close, the cash payments involved might be only a few ten-thousandths of the notional contract value.

Of course, as we all discovered in the past couple years, sometimes the unexpected happens, and in that case those contractual payments can blow up, leading to a situation where company X owes company Y a significant fraction of all the money in the world, so company X goes bankrupt or gets a bailout.

[russell]

This is very misleading, even if technically correct.

Quite possible. It's not my field, it's just the fishbowl I'm required to live in, so I'm trying to make sense of it.

Of course, as we all discovered in the past couple years, sometimes the unexpected happens.

OK, then maybe not so misleading?

I guess my question is whether financial markets leveraged at 30 TIMES THE PRODUCTIVE OUTPUT OF THE ENTIRE FREAKING WORLD is something that should get our attention.

By "leveraged" I mean the thing where the cash payments are only a few ten-thousandths of the "notional" contract value. Because when the rubber meets the road, the value is not just "notional".

And by "get our attention" I mean, "do something about".

If this is a topic you are well informed about, I invite your further thoughts.

[alkali]

I hear you but this is not a very informative statistic. By way of comparison:

1) The total of all homeowners insurance premiums in Texas in 2008 was about $5 billion. The total face amount of those policies is likely upwards of $1 trillion.

2) The total amount of new construction in Texas in 2008 was about $50 billion. Assume (charitably) that half of that -- $25 billion -- was residential homes.

3) So the total value of homeowners insurance written in Texas in 2008 is at least 40 times the amount of new home construction in Texas in that year, and probably much higher. That's comparable to the statistic you generated.

Doing this kind of calculation is not particularly informative because you are comparing a stock variable (insured amount) with a flow variable (output per year).

[Tony P.]

Russell,

Aside from Peter's point about the difference between notional value and cash at risk, there's also the difference between "turnover" and aggregate value. It seems (but I could be wrong) that if a "betting slip" worth a billion dollars gets bought and sold a thousand times, it contributes a trillion dollars to the turnover number.

I suspect that, in a similar way, one could add up the number of shares traded on the NYSE in a given year and come up with a total that's ten times the number of shares of stock outstanding in the entire world.

But you are of course correct to point out how insanely much "business activity" consists of furiously trading betting slips back and forth. Add in the fact that the traders, like the house in Vegas, always end up with a cut of the turnover, and it's enough to justify pitchforks and torches.

--TP

[russell]

I'm going to assume that the homeowners insurance policies in TX cover more than just new construction, so to me it makes some sense (to me) that the total face value of all homeowner's policies is 40 times new construction.

Annual premiums at one half of one percent of insured value seems low to me at a sort of knucklehead intuitive level, but I'm going to assume that there is an actuarial basis for the numbers.

In other words, somebody has sat down and run the historical data and figured out that the odds of damage to a home in TX that will result in a payable claim is something like 5 in 1,000.

I'm not so sure that same due diligence has been done with derivatives. I could be wrong, and maybe the events of the last few years are just a run of really, really, really bad luck.

Odds are that they are not. That's a lot of bad luck.

In addition, I'm thinking the TX home insurance market is not going to sell me, a resident of MA, an insurance policy on a home in TX that I do not own and have no financial interest in. Unlike, frex, the derivatives market.

Maybe the two markets are kind of equal, but you haven't really won me over. Feel free to expand your comment if you like.

[Edmund in Tokyo]

Is the $1,200 trillion number for the amount somebody is liable for, or is it the amount that's gone backwards and forwards during the year? I read it as the latter, so:

- If you sell me an insurance policy for my cat for $10 paying out $50 if my cat dies, the total traded is $10, with a liability of $50.

- If you sell that liability to your friend for $10, then they sell it on to their friend for $10, and keep going 1000 times, the total traded is $10,000, but the liability is still only $50.

- If the liability then gets sold back to me, the total traded is $10,010, but effectively the liability is $0 and I have an uninsured cat.

[jdkbrown]

According to this article:

http://dollardaze.org/blog/?post_id=00565

there were roughly $58.9 trillion in the M3 money supply in 2009. The M3 money supply is: all bank notes and coins in circulation or in reserve + all checking accounts + all small savings and term deposits + "large time deposits, institutional money market funds, short-term repurchase agreements, along with other liquid assets".

If I'm reading the article right, this is, essentially, the value of all the money in the world--a stock variable--and not a flow variable. So it seems russell's point is right: the market in derivatives was twenty times the value of all the money in the world.

[russell]

there's also the difference between "turnover" and aggregate value.

That's a good point. And, if the number we're talking about is just the nominal value of all of the trades added up over the course of the year, as opposed to the total face value of all of the instruments themselves, I'd say that freaks me out somewhat less.

I'd be curious to know total face value of the derivative products (as opposed to adding up that face value per trade) as a function of GDP. I'll see if I can track that down.

Net/net, what strikes me is that the function of the financial sector has moved from finding productive uses of capital, to some kind of weird self-referential wankery.

I understand that there is some value in derivative products as hedges, but when the face value of the hedge is many multiples of the thing it's hedging, it seems kinda problematic.

Like, if the total face value of all homeowner's insurance policies in TX was 40 times the value of, not just new construction, but all residential housing in TX, that would be weird.

Like, somebody-should-go-to-jail weird.

Maybe I'm missing something....

[alkali]

My point isn't that the markets are equal -- I don't think they are. My point is that the calculation that you are doing is not a reliable way to tell whether too much of an insurance product has been issued.

I'm not so sure that same due diligence has been done with derivatives [as with homeowner's insurance].

Agreed that that's the key issue.

[ral]

"Leverage" and "turnover" -- music to a broker's ears.

"Sometimes the unexpected happens." Like LTCM for example. Really, is this still surprising? Only the scale, not the fundamentals.

[russell]

The BIS has the notional (not turnover) value of OTC derivatives at $614 trillion as of December 2009.

Market value is $21 trillion, by which I think they mean "what it would cost you to buy all of those instruments".

The relationship of "face value" to "what you're on the hook for if the sh*t hits the fan" is not completely clear, to me. Although I suspect the folks holding those instruments are, in the aggregate, on the hook for something like the $614T.

Global GDP in 2009 was about $61 trillion.

So, face value of OTC derivates to the productive output of the entire world in 2009, ten to one.

Maybe, like alkali's TX homeowner's insurance market, that's kind of OK.

Then again, maybe it's not.

Note also that the OTC derivative market is largely unregulated, and mostly consists of agreements negotiated directly between one party and another. IIUC.

[ajay]

http://en.wikipedia.org/wiki/Notional_value

HTH.

russell: the markets weren't "leveraged at 30 TIMES THE PRODUCTIVE OUTPUT OF THE ENTIRE FREAKING WORLD". Leverage is something completely different.

[jpe]

Those numbers are gross notional, which don't tell us much. Example: say you and I bet on a football game, and say the bet will be based on a notional one million dollar contract. The loser will pay the winner .001% of the contract value, or $10. So we've got a notional one million dollar contract and ten bucks will change hands. Is our bet ten bucks or a million bucks?

That's the difference between notional value and actual value.

Say I don't like my bet anymore, so I bet the opposite way w/ another person so that no matter what happens I won't win and I won't lose. Now there's notional two million dollars out there, but still just ten bucks will ultimately change hands. What's out there? Two ten dollar bets or one? So that's the difference between gross notional and net notional.

You're just looking at gross notional, which doesn't tell us anything meaningful about potential losses or the total value of the contracts.

[jpe]

The relationship of "face value" to "what you're on the hook for if the sh*t hits the fan" is not completely clear, to me.

It depends on the type of contract. W/ a credit derivative, the seller is generally on the hook for the full face value. With the other variants, the notional value is never at stake, and just serves as a fictional reference point for other calculations. You can take net notional credit default swaps and say that that's the total that's on the hook, but CDSs are something like 5-10% of the total notional number last time I checked.

[wj]

I think it helps to see two kinds of derivatives. The first (and original) kind are basically insurance. One party is insuring against changes in the price of something. The other is selling insurance against such a price change. And, like any insurance company, setting the price of that insurance to cover the inevitable cases where it has to pay off.

The other kind, representing the vast majority of the derivatives we are discussing, are straight bets. With the selling party simply not having the resources to cover their bets.

There's a reason why most casinos and other betting establishments favor "table stakes" -- it makes sure that any bet made can be covered. Any sensible regulation of the derivatives markets would do the same. (And, no doubt, be denounced as an outrageous interference in the free market.)

[ral]

The wisdom of Tom Lehrer:

Be prepared
To hide that pack of cigarettes
Don't make book
If you cannot cover bets

[Bruce Webb]

(Well I see now that jpe made the same basic point above. Call this reinforcement.)

If I bet $2 on a $1 million dollar race horse running in a major stakes race and the horse wins, between the purse and the increased value of the horse for future breeding the total value of that horse might go up by 5% or in this case $50,000 or more. That doesn't mean I am making either a $1 million dollar bet or even a $50,000 bet, it is still a $2 bet that might (if the horse was an odds on favorite) only return say $2.75. And if I made that same $2 bet on a 1st race $2000 claimer that went off at the same odds and wins I get the same return. That is the value of the bet and even the entire pari-mutual pool may have zero correlation with the asset(s) whose performance is drawing the bet.

[Slartibartfast]

I'm still not sure if russell is on to something or if he's made a mistake.

If I trade my 1970 Plymouth Superbird for an equivalently priced but different classic automobile, the one-directional value of that trade is (just for example) about $100k. $200k worth of assets has just moved. Now, if I continue trading approximately like value for like value, I could wind up with a 1930 Rolls-Royce Phantom that suits my elderly, more educated but spectacularly obese buttocks better. Suppose it takes me ten car swaps to get there, then $2 million in assets have moved, but net change in asset value as held by me and everyone I've traded with is approximately zero.

I'm wondering if the 1.2 quadrillion is more like that.

It'd be nice to hear from more people who aren't doing the blind men/elephants thing with this.

[Sebastian]

This is one of those things where I agree with everyone above and want to make a further point.

As above, the notional value is close to useless for analyzing how serious the issue is. An additional reason why that particular number is useless is that they are essentially double counting--both sides of the bet are being used for that number.

Now I'm perfectly willing to believe that derivatives can be a big problem. Any leverage vehicle can be a big problem. But the problem is the extreme leveraging, not any particular 'riskiness' of derivatives in general.

[Thermopylae]

Sensational post.

---
The word “bet” applied to derivatives in this context is being used disingenuously. Accordingly, people who purchase insurance (car, life, et al.) are essentially “betting” and “speculating.”

---

The size referenced of the derivatives market is a turnover figure.

So, if a trader day-trades $1B contracts that are backed by 100% collateral (that is, $1B) everyday for one year, then doing so will be recorded as $250B for the year (50 weeks).

In other words, the turnover would be 250x the collateral value. However, in reality, this trade was supported by 100% collateral.

[ral]

The problem leading to high risk in derivatives is on the selling side. Consider simple "puts" and "calls" on a stock.

When you sell a put you promise, for a limited duration, to accept shares at a specified price, no matter the market price at the time. So, if the market price goes to 0 your risk is the face value (the "striking price" times the number of shares).

When you sell a call you promise to deliver shares at a specified price, again regardless of the market price. In principle your risk is unlimited since the market price might skyrocket.

The way to insure against the problem is to have a stock position that backs your promise. That is, if you already own the shares you sold the call on you can simply deliver them. If you write a "naked" option (not covered by a position), you are taking a risk. AIG effectively wrote naked options.

[russell]

I'm still not sure if russell is on to something or if he's made a mistake.

Me either.

Or, I was certainly mistaken in thinking that the notional value of the derivatives market is an indication of the risk exposure.

I'm still trying to get a reasonable understanding of what the risk exposure actually is. It's hard to do that because a lot of these instruments are essentially private deals between a couple of parties.

Sensational post.

Not intentionally. More a combination of "insufficiently informed" and "freaked out", the latter being some function of the former.

Just trying to figure out what's what.

The word “bet” applied to derivatives in this context is being used disingenuously.

Not so sure.

From my new and improved understanding of derivatives, I would say that many fall into the "insurance" category. They are used as hedges, to limit exposure to risk.

Lots of others appear to be, straight up, bets.

It'd be nice to hear from more people who aren't doing the blind men/elephants thing with this.

Agreed.

[Countme?]

The Standard and Poor's 500 has gained next to nothing annually since 2000, and it was at much higher levels three years ago, and yet ..........

http://articles.moneycentral.msn.com/Investing/Dispatch/default.aspx?feat=1814955&GT1=33009

As they've always asked, where are the customer's yachts?

[hairshirthedonist]

From my new and improved understanding of derivatives, I would say that many fall into the "insurance" category. They are used as hedges, to limit exposure to risk.

Lots of others appear to be, straight up, bets.

Yes. I'd say that my homeowner's insurance is not a bet, because I am required to cover it by the terms of my mortgage, because I live in and partly own it, and because I cannot afford to replace my house if it burns to the ground.

For russell to take out insurance on my house, in which he otherwise has no direct or indirect interest, hoping it to burn down, would be a bet.

[Bernard Yomtov]

From my new and improved understanding of derivatives, I would say that many fall into the "insurance" category. They are used as hedges, to limit exposure to risk.

Lots of others appear to be, straight up, bets.

The problem is that the insurer, usually, is making a bet.

Your homeowner's insurance company is betting that your house won't burn down. The only gain or loss to them is financial.

Take the standard example, suppose you're a farmer who wants to hedge (insure) the price of the wheat you're going to harvest six months from now. You can enter into a futures contract, agreeing to sell the wheat in six months but at a price specified today.

But someone has to be on the other side of that; you need what's called a counterparty. It would be nice if there were just the right number of wheat buyers - food companies, say - to match up delivery times, prices, etc. with all those farmers. But there aren't. Enter speculator, stage right, twirling moustache. The speculator takes the other side, betting that the price of wheat will rise, and giving him a profit.

And of course the food companies will often deal with speculators also. Without these evil creatures the system wouldn't function.

Of course the speculator would really like to deal with both sides, at slightly different prices, just as a bookmaker bets on both teams at slightly different odds. And many custom derivatives work more or less that way. The dealer designs something to hedge a risk and then tries to act as an intermediary, taking a spread out of it. And this can happen with standardized contracts, such as some kinds of swaps, as well.

It's critical to understand the role of the "market-maker," because that's what these people's role is. They provide liquidity to the market, in exchange for earning a spread between what they buy and sell for.

In a way, this is no different than a used car dealer. The dealer will buy your car, for a dismal price, and try to sell it at a profit. The service he provides the seller is saving the time, expense, and effort involved in selling the car, and providing instant cash. He's a market-maker in used cars.

[wj]

Bernard, you can call what an insurer does a "bet," but it differs from what I (for one) call a bet in that the insurer sets his prices in the knowledge that he will pay on some of the policies he writes. And sets aside assets accordingly. His decision on how much to set aside (or, if you prefer, on how many he will have to pay on) is based on an analysis of statistics on a large number of similar contracts in similar circumstances.

A speculator making a real bet, in contrast, is working on hope. At best, he has a "gut feel" about how prices might move. But does someone speculating in hog belly futures actually know anything about agriculture, or about market demand for bacon? Mostly, no.

A market maker is, as you show, in a different position. He is taking a fee (the spread) for joining up a seller (who is hedging his risk of a price drop) with a buyer (who is hedging his risk of a price rise). Other than those who are in it to place bets, the sellers the market maker works with actually expect to have the product to deliver on the due date; all they are hedging is price movements. And the buyers expect to actually use it. Unlike the speculators, who don't expect to have the product (although they may have to buy it to fulfill their contract if they lose their bet) and don't have any use for it (even though they may have to pay for it if that bet goes bad).

[Bernard Yomtov]

wj,

The insurer is making a bet the same way a casino does, with the odds in its favor and with enough action to pretty much assure that the outcome will be profitable. But it's still a bet - that is, an agreement to pay money based on some random event.

A speculator is not necessarily operating on gut feel, but may well be in the same position, albeit less securely, as an insurer. Remember that the hedger is risk-averse, and will often agree to a futures price less than the expected actual price (the expected spot price). That is called "normal backwardation," IIRC. In such a situation the speculator, like the insurer, is making a bet at favorable odds.

But even if that's not the case, the main point remains - the hedger needs the speculator.

And even the market maker often has to speculate somewhat, since risks will not perfectly offset one another.

[Tom M]

The insurer is making a bet the same way a casino does, with the odds in its favor and with enough action to pretty much assure that the outcome will be profitable. But it's still a bet
Not at all. The casino is playing a game with a well known set of variables. The odds of an outcome outside its experience is zero. It may not be able to handle all the possible outcomes but it knows what they are.
The insurer is using history as a guide to the future, it's an experienced type bet but history is not the same set of experiences as playing a game. What happened in the sub-prime mortgage market is the exemplar of the difference between history as a guide and history as a limit to the actual outcomes.
The models behind the RMBS and resultant CDOs etc. predicted default rates based on historical outcomes, including extending credit on autos, credit cards etc. The historical default rates were the baseline.
But the outcomes were not at all like those in a casino. It would be like saying the casino in craps has to pay based on the expected incidence of two sixes unless two sixes came up three times in a row and then the casino would have to pay ten times the usual amount.
No casino would take that bet but that is precisely what happened in the RMBS and especially in the CDOs. Which is why the notional amount of the bets really does matter.
You only think you know all the possible outcomes and their effects. But you have recent massive empirical evidence that you don't. And yet you liken it to a casino?

[hairshirthedonist]

I knew this would turn into an argument over the word "bet" at the expense of the clear and real distinction being made. I think the key is russell's use of the words "straight up" in describing pure speculation that doesn't involve some sort of protection of an actual interest. The word "bet" is simply a vehicle or a point of focus for making that distinction. I don't think anyone is trying to re-write the dictionary or to suggest that it's absurd to use the word "bet" in describing homeowner's insurance in any and all contexts.

[ruidh]

I think it's incorrect to characterize these kinds of trades as "betting". The goal for an investment bank which wants to be an intermediary in a swap contract is to find someone to take the other half so that the bank becomes merely an intermediary passing the net proceeds back and forth and perhaps taking a small cut. For each person betting that the floating rate will go up, there's someone else betting it won't go up that much. If they can't find someone to take the other half, they have to go long or short on Treasuries to hedge themselves.

[wj]

Bernard, a casino does not make bets, it takes bets. Its patrons are the ones who make bets.

The casino sets the odds for each game (or selects games for the odds, if you prefer) to ensure that it will make money overall. (Note the similar roots of insure and ensure.)

[russell]

I think it's incorrect to characterize these kinds of trades as "betting".

I think what you're saying is true for pure market-making activity, where the bank engages in offsetting trades that reduce it's own risk.

There is also a large amount of derivative trading that is purely speculative, and that seeks to generate income by coming out on top of the trade, rather than just get paid for creating the market.

If the term "speculative trade" is preferable to "bet", that's OK with me, but I don't really see a lot of daylight between the two.

[hairshirthedonist]

Neither I nor my home's insurer want my home to burn down. Nor does my mortgage company. Compare that situation with a certain number of credit-default swaps.

[Tony P.]

Betting is the archetypal example of a ZERO-SUM GAME. What one bettor wins, the other loses. To the extent that derivatives trading is NOT a zero-sum game, then to that extent it is not "betting".

I think a good case can be made that the Great Panic of 2008 proved beyond a doubt that derivatives trading was not a zero-sum game. It was a NEGATIVE-SUM game. It was like a poker game where ALL THE PLAYERS LOSE.

If you think I'm kidding, please explain to me who it was that the players in the derivatives game lost their money TO. Maybe I did not pay close attention, but the financial panic was NOT described to us as "OMG!! These investment banks over here lost all their money to those hedge funds over there!" It was NOT described as half the financial system winning trillions of dollars from the other half. The financial system, we were told, had turned into a negative-sum game.

So, no: all those Troubled Assets that needed a Relief Program were not "bets", because a bet implies a winner as well as a loser.

--TP

[Bernard Yomtov]

HSH,

I knew this would turn into an argument over the word "bet" at the expense of the clear and real distinction being made. I think the key is russell's use of the words "straight up" in describing pure speculation that doesn't involve some sort of protection of an actual interest. The word "bet" is simply a vehicle or a point of focus for making that distinction. I don't think anyone is trying to re-write the dictionary or to suggest that it's absurd to use the word "bet" in describing homeowner's insurance in any and all contexts.

Yes. While I disagree with wj over the definition of a "bet," there's no need to get into an argument over that.

Russell,

There is also a large amount of derivative trading that is purely speculative, and that seeks to generate income by coming out on top of the trade,...

Yes, but to the extent the speculator is assuming the hedger's risk this is useful. Again, the speculator, like the insurance company, is willing to take a risk in exchange for a premium. That premium may be favorable odds, or a direct payment. Note that in the options markets the prices of options are even called "premiums," just like the price of insurance.

A useful distinction between this sort of speculation and gambling is that the speculator assumes unavoidable risk, while the gambler creates risk.

For example, currency exchange rates are going to fluctuate. If you are in a business that is affected by those rates the risk is unavoidable. But what you can do, with derivatives, is pay someone, possibly a speculator, to assume the risk, just as you pay the insurance company to assume the (financial) risk of your house burning down.

OTOH, there's no inherent reason the outcome of a football game should cost you or me any money. If we bet on the game we have created risk out of thin air. We are gambling.

Tony P.,

Betting is the archetypal example of a ZERO-SUM GAME. What one bettor wins, the other loses. To the extent that derivatives trading is NOT a zero-sum game, then to that extent it is not "betting".

But you need to be careful to specify in what terms it is zero-sum. Insurance is zero-sum in terms of money, but not utility. Why do we buy insurance (assuming we're not required to by our mortgages)? The reason is that we are risk-averse. We're willing to lose a little bit - the premium - to avoid a catastrophic loss. We are willing to pay a $1000 annual premium, say, to avoid losing $100,000 even if the chance we will lose the $100,000 is less than 1%.

The insurer, by contrast, is risk-neutral. It's looking only at the odds, so it is happy to to assume a .6% chance, say, of a $100K loss in exchange for $1000. Why? Because it has a lot of money, first of all, so the loss does not have consequences beyond the payout. And more important, because, like the casino, it is taking lots and lots of presumably independent risks of this type, so it can quite reasonably expect to come out ahead by $400 per policy.

It is this difference between risk-aversion and risk-neutrality that makes insurance a positive-sum transaction on a utility basis, even though it is financially zero-sum (or worse, once you count administartive costs).

[wj]

Tony, suppose that you live in a low-lying area and therefore buy flood insurance. And, for whatever reason, do not buy enough of it to actually cover the value of your house and all of your possessions. Now, before you have been paying premiums very long, there is a flood. The insurance company loses - they have to pay out far more than they have made in premiums. You lose: you do not have as much value as you had before the flood. So, you bet any flood would not be too bad, and you lost; the insurance company bet there wouldn't be a big flood any time soon, and it lost. Who won there? Mother Nature, perhaps?

In the case of derivatives, those holding them when the music stopped lost. And those who wrote them lost, too. The only people who "won" were those who bought the underlying assets, created derivatives from them, sold the derivatives for more than the original assets cost, and got out without holding many themselves.

[wj]

Hmmm, rather than "those who wrote them" what I intended to say was "those who held the underlying assets" (e.g. the house to which the original mortgage was attached).

[ral]

No, wj, there were some other winners. Large winners.

[Jonathan Weber]

"-- If you think I'm kidding, please explain to me who it was that the players in the derivatives game lost their money TO. Maybe I did not pay close attention, but the financial panic was NOT described to us as "OMG!! These investment banks over here lost all their money to those hedge funds over there!" --"

That's because the financial crisis was the result of a devaluation of property, not the shuffling of money.

If a house is worth $500k today and $300k a year from now, what happened to that $200k difference? Who "got" the money? No one. In this situation, you have banks issuing $500k mortgages and foreclosing on $300k houses. The $200k difference in price didn't "go" anywhere. People just aren't willing to pay the sale market value for the house anymore.

It's the same way that your car devalues when you drive it off the lot. You start with $10k and the dealership starts with a $10k car. When the sale is complete, the dealership has $10k and you drive off with an $8k car. The economy shrinks by $2k.

Likewise, if a bank issues $10k in debt and you pay back $8k before saying, "Nope - I don't have any more!" and default, that's $2k no one has anymore.

[russell]

That's because the financial crisis was the result of a devaluation of property, not the shuffling of money.

I think you underestimate the contribution that financial speculation made to the devaluation of property. Or, rather, to the artificial inflation of its value.

The word is "bubble".

To clarify both my choice of words here, and my concern:

When say "a bet", all I mean is that somebody is putting money at risk in the hope of a larger return, and that return depends on circumstances playing out in a way that they cannot reliably control or predict.

I suppose you could consider insurance "a bet" in that sense, but IMO the actuarial discipline behind most insurance products kind of puts it in another camp.

Bets, in the sense I'm using the word, can be well-informed or bone-headed, but the basic mechanics of the transaction are the same.

I most definitely recognize that financial risk-taking has an upside, and I also recognize the value that derivatives provide as way to shift risk from folks who would prefer not to have it, to folks who don't mind having it at all. I'm not looking to eliminate derivative products from our economy.

My concern is the degree to which derivative products are the vehicle for really dangerous amounts of leverage. Also, the way in which, in practice, they've had the effect not of insulating sectors of the economy from risk, but of spreading the damage quite broadly.

Last but not least, I'm concerned about the way that investment banking - which used to be primarily private companies and partnerships, where people put their own money at risk - has become an environment where bankers take very large risks with other people's money, and where, in the context of publicly traded companies, they reward themselves with *very* high levels of compensation.

My understanding is that the GS executive who took Ford public in 1956 got paid $250K for putting the deal together.

Things have changed.

[Marty]

"My understanding is that the GS executive who took Ford public in 1956 got paid $250K for putting the deal together."

What was the deal value, because thats about 2M in todays dollars, a solid bonus for a small deal even today.

[russell]

It was the largest IPO ever done up to that time. It took about two years to put together.

I'll see if I can track down the dollar amount.

[Marty]

russell,

Rhetorical question. It would be more interesting to know how much the bonus for the executive responsible for the Google IPO was, not asking you to do homework, it is my way of wondering if things have really changed that much:

[russell]

10.2 million shares, $660 million in 1956 dollars. About $5.3 billion in today's money.

[russell]

Hey Marty -

Don't know about Google specifically.

Per this article, circa 2006 US IPO fees ran about 7%, UK IPO fees about 4%.

7% of $660M is about $46M. 4% of that number is $26.4M.

Note to self: if I ever IPO russell.com, I'm getting a UK underwriter.

[Marty]

russell, yep 7% is about right for the firm, negotiable but a reasonable average. I asked a few people I know what the bonus for the lead banker would be out ofthat. I will let you know if I get an answer.

[Marty]

BTW, spend the extra 3% on GS, they are worth it to you.

[russell]

you got a point there. big picture, 3 points is short money.

[Tom M]

My concern is the degree to which derivative products are the vehicle for really dangerous amounts of leverage. Also, the way in which, in practice, they've had the effect not of insulating sectors of the economy from risk, but of spreading the damage quite broadly.
That's pretty succinctly how BearStearns, Lehman, and Merrill went out of business. Bear was leveraged about 40 to 1 and when they lost their "bet" on subprime mortgages, it didn't take a lot to make them insolvent.
Those CDOs and RMBS fails did pay off and somebody did win. See Michael Lewis' The Big Short for a depiction of a few of them and how they did it.

[Tony P.]

If a house is worth $500k today and $300k a year from now, what happened to that $200k difference? Who "got" the money?

Well, let's stipulate that the house did not shrink by a couple of rooms or get 40% burned down between last year, when I paid you $500K for it, and this year, when nobody will pay me more than $300K for it. It's the same house, right?

Now the answer to who got the missing $200K is pretty obvious: YOU did. For you could buy the house back from me today and still have $200K of my money left in your bank account.

It's still the same house, remember. As a piece of physical property, it has the same value it always did. The net result of the price drop was not a decrease in the agreggate REAL wealth of the nation. It was merely a redistribution of money from me to you.

So far, the story is terribly simple. The price of a house (or an asset in general) is determined by nothing except what some buyer pays some seller for it. The buyer's money becomes the seller's money. If the buyer "lost money", it was the seller who "got" it.

Now consider the identical house next door. It was not bought or sold in either year. But its notional price is of course always the same as the price of our house. So its owner ALSO "lost" $200K, right? Where did HER $200K go?

I suggest it never went anywhere because she never had it in the first place. Before the price bubble she was living in a 3BR house "worth" $300K; after the price bubble she's still living in the same 3BR house that's still worth $300K.

Imagine, however, that at the peak of the bubble her son the financial wizard said, "Mom, you're living in a paid-up house you bought for $300K. The house next door just sold for $500K. You don't want to sell your house, but why don't you take out a $500K home-equity loan." So the old lady borrows $500K of Other People's Money and spends it on that year-long luxury cruise around the world she always dreamed of. She figures that when she gets back she can sell the house, pay off the loan, and move into assisted living.

Alas, the house will only fetch $300K now. So her son the wiz says, "Mom, only a sucker would keep making payments on an underwater mortgage. Just mail the keys to the bank and walk away." Net result: who lost $200K in this case is the Other People. In essence, THEY paid the old lady $500K for a $300K house.

Had the old lady sold her house for $500K to spend on a cruise, last year, the buyer would be the only loser now. But she did not. And why not? Because she "speculated" that the price of the house would go UP, that's why. Her speculation saddled those Other People with $200K worth of her half-million-dollar cruise. SHE got the $200K they lost.

Had the old lady just held on to the house right through the price bubble, she would neither have "made" $200K in the run-up, nor "lost" $200K in the crash. She would not have got her cruise, either, of course.

Had the old lady sold at the peak of the bubble, SHE would have "got" the $200K that the buyer "lost". The BUYER would have paid for 40% of her cruise.

What cost those Other People $200K was "speculation" and "finance": the old lady's speculation that the price of her house would go up (or at least, not down); and the Other People's financing of her speculation. But there's still no mystery about who "got" the money: the cruise company got it.

--TP

[RogueDem]

TP, I think you have it backwards, the party making the loan are the speculators not the one taking out the loan. The Other People were the ones speculating that the house would increase in price while the little old lady essentially sold her house to them for 500 G’s. The Home Equity Loan was collateralized with her house, she got the loan and eventually gave up the house. When the loan was made, there were two possible outcomes: 1) she keeps making payments and for that the bank lets her live in the house until the loan is paid off, or 2) she stops making the payments and the house goes over to the bank. Both parties knew the possibilities going in, it was the bank that was speculating and ended up with an asset that they had overvalued.

[Tony P.]

RogueDem,

You make a good point, to which I respond that it takes two to speculate.

I said the old lady "speculated" that the price of her house would go up. That was not totally wrong. She held on to the house in hopes that its price would go up. That's pretty much the definition of what speculators do. But I recognize that some people hesitate to call it speculation when somebody merely holds on to an "asset".

The Other People (or the bank, if you prefer) "speculated" too, as you say. In "financing" the old lady's round-the-world cruise, they bought an asset (her note) in hopes of a return. Some people might hesitate to call it speculation when somebody merely lends money at interest, secured by collateral. But you don't, and neither do I.

--TP

[LongHairedWeirdo]

And of course the food companies will often deal with speculators also. Without these evil creatures the system wouldn't function.

Speculation is a particular type of activity - high risk gambling that could pay off extremely well. You don't need speculators. (There are those who argue that speculators serve a purpose - drawing attention to possible fluctuations. But that's not *necessity*.)

You do frequently need people who will "gamble" on these contracts, because businesses are too careful with their money to gamble this way. These contracts give people the ability to plan for the future. But you don't need speculators.

[LongHairedWeirdo]

I said the old lady "speculated" that the price of her house would go up. That was not totally wrong. She held on to the house in hopes that its price would go up. That's pretty much the definition of what speculators do. But I recognize that some people hesitate to call it speculation when somebody merely holds on to an "asset".

The old lady did not engage in any form of speculation. She did gamble in a manner that others were swearing was okay.

The problem in this case was that the bank should have recognized that it's risky to lend that much money on a house to a person who probably can't afford to pay it, and must pay it off by selling it.

But some financial wizards at the bank said "it's no risk at all! We'll just bundle up the mortgage as bonds, and sell the risk off! People are buying these bonds like hotcakes!"

Since there seemed to be no risk to the bank, and pure profit generated by making the loan, the bank wasn't speculating either.

The funny thing is, the buyers of the loans were all following conventional wisdom, that prices were rising, and people just don't *walk away* from mortgages, so they weren't speculating either.

At some point, it should have been obvious to everyone that the bubble was a bubble, and unsustainable. The problem was, it was now a game of hot potato. How do you cover your own positions, and avoid an ugly bankruptcy?

Here, buying and selling would have turned into speculation, except there was a secondary crisis.

Everyone was leveraged. They'd borrowed money to buy these mortgage bonds, based upon the value of the bonds. And they had to mark to market - those bonds were worth as much as they could be sold for, no more. And no one would buy them any more.

If they marked down the price low enough to find a buyer, then their loan holder would say "you don't have enough assets to cover the loan, come up with more, or pay off the loan." (This is known as a margin call.)

Speculation didn't happen, because no one could afford to sell at prices where speculators started to nibble - the margin calls would make them collapse like a house of cards.

So, what happened wasn't the result of speculation. It was the result of a lot of money games, where everyone tried to act in their own best interests first (and their company's second), and piled up a lot of imaginary assets.

Sometimes this works out well - a hot startup company needs capital, it starts selling stock at $10, people think it'll be worth $50 next year, and buy it at $40. Suddenly, that company has $30/share of purely imaginary money - and, if it's run well, might just be worth an honest-to-god $50/share next year.

But sometimes, that imaginary money is based upon an unsustainable set of price increases in houses, magnified and multiplied by nothing but blind confidence that "the system" won't collapse... until it does.