Anyone around here either physically or online will have heard of the famous Mischel Marshmallow experiments that examined the patience of children, measured by their ability to delay consumption of a marshmallow in exchange for getting two marshmallows later on. We tried to replicate them here but we kept eating the marshmallows before we could recruit the participants.
Nudge blog has a description of the famous experiments along with some video – link here
Via Colm, below is a link to a recent conference at Melbourne on the economic situation in Australia. I am listening to the recordings in the background as I work here. This is a very well themed conference in terms of the balance between short-run and long-run policy. Includes sessions on innovation, education and getting people back to the work.
As I’ve said before, I think the idea that “too big to fail, too big to exist” idea is just silly—it betrays a fundamental lack of knowledge about the way modern financial markets work. The pundits who push this idea love to argue that banks don’t need to have huge balance sheets, and that there’s no benefit to having banks with balance sheets of over, say, $400 billion or so. This argument, too, is almost adorably naïve. (The whole thing can also be refuted in four words: Long-Term Capital Management.)
It’s been odd to watch the debate on bank size though, because the people defending big banks in the media/blogosphere (e.g., Charles Calomiris) have somehow managed to avoid mentioning the one reason banks do actually need very large balance sheets: market-making.
The major banks are all market-makers (or “dealers”) in fixed-income products, currencies, OTC derivatives, commodities, and equities. In general, dealers in a given security stand ready and willing to buy or sell the security for their own account, at publicly quoted bid and offer prices. Market-making, especially in fixed-income products, is very capital-intensive.
You need a very large and diverse balance sheet to be a market-maker in fixed-income products—government securities, investment grade corporate bonds, high-yield bonds, mortgage-backed securities, bank and secured loans, consumer ABS, distressed debt, emerging market bonds, etc. Dealers hold inventories of all these securities because they need to remain “ready and willing” to sell, and because when they buy a security from a client, they need to hold it in inventory until a buyer for the security appears. Dealers are exposed to price movements for the period they hold the security in inventory, and because inventories can grow large in a short amount of time, sharp price movements can result in substantial losses for dealers.
So dealers hedge. Constantly. The cheapest way for dealers to hedge is internally—that is, when the security or derivative it buys can offset an exposure elsewhere on its balance sheet. The next cheapest way for a dealer to hedge is generally with liquid, vanilla derivatives (e.g., interest rate swaps). So imagine an MBS dealer that buys a large position from a client, and has to hedge the interest rate risk. If the dealer also happens to be a market-maker in interest rate derivatives, then either: (a) the interest rate risk on the MBS will offset one of the rates desk’s exposures; or (b) the rates desk will go into the market and hedge the interest rate risk with a plain-vanilla derivative, which it can do very cheaply because as a market-maker, it does these kinds of trades all the time. So being a market-maker in interest rate derivatives is critical to effectively managing the risks of holding MBS in inventory—and if you can’t effectively manage the risks in a large inventory of MBS, then you simply can’t offer cost-effective market making in MBS. What’s more, dealers also need to set aside capital for their market-making in the OTC derivatives that they use to hedge their fixed-income market-making.
Now think about all the different kinds of risks involved in holding inventories of the fixed-income products I listed above. We’re talking about foreign exchange risk, interest rate risk, credit risk, basis risk, etc. Hedging all of that, dynamically, is a necessary component of market-making. This is why, for example, Goldman bought CDS protection from AIG on the super-senior tranches of CDOs it underwrote. The point of creating CDOs was to generate a mezzanine tranche, which investors, who had a seemingly insatiable thirst for yield, would gobble up. Goldman (and other dealers) couldn’t place the super-senior tranches, so they held the super-seniors on their books and hedged all that risk by buying CDS protection from AIG (and the monolines). There’s nothing nefarious about this—hedging is just what dealers do. Alas, this concept is apparently too difficult for the Matt Taibbis of the world to get their minds around.
As you can imagine, all the risks that a major dealer bank has to manage on a daily basis—the constantly changing level of their exposures, how those exposures all interact, etc.—gets extraordinarily, mind-bogglingly complicated. The major banks all made huge investments to develop the technological capacity to manage those risks, and it’s pretty clear they didn’t invest enough in their risk management systems. There are only two banks that I’ve seen that clearly did make the necessary investments in risk management (Goldman and JPMorgan, not surprisingly). So there are undoubtedly economies of scale there.
Another place there are economies of scale is order flow. The larger a dealer’s order flow, the more trades it can match internally. This reduces volatility, allows a dealer to hold smaller inventories of securities, and reduces its exposure to sharp price movements. A lot of the financial industry’s “merger mania” over the past 15 years was driven by the race to capture order flow.
So why do we need these massive market-makers in the first place? They ensure liquidity in the capital markets. And why is that important? For one thing, it lowers borrowing costs—investors are much more willing to buy a bond issue if they know they can quickly and easily sell the position later if they want to. Investors demand higher yields for illiquid bonds. The benefits of having massive market-makers were passed on to all the businesses that were able to borrow in the capital markets at a much lower cost, and to all the investors who enjoyed much higher returns due to the reduced transaction costs.
Having liquid capital markets also allows the use of mark-to-market accounting, which is an important check on corporate management. During the whole nationalization debate, everyone was screaming bloody murder about the fact that the banks didn’t have to mark their toxic assets to market. Well, if we “break up” the major banks, as some simpletons pundits are urging, then you can forget about being able to mark-to-market lots of fixed-income products and OTC derivatives.
Of course, there’s no chance that we’re going to break up the major banks. Tim Geithner isn’t an idiot, and he’s not an attention-craving pundit. Unfortunately, most pundits are apparently unable to distinguish between recognizing the benefits of big banks and being “captured by Wall Street” (which is a red herring). So if you’ve somehow made to the end of this post, I hope you’ll be able to see through the claims that Geithner’s unwillingness to break up the major banks is proof that he’s been “captured by Wall Street.”
Anyone around here either physically or online will have heard of the famous Mischel Marshmallow experiments that examined the patience of children, measured by their ability to delay consumption of a marshmallow in exchange for getting two marshmallows later on. We tried to replicate them here but we kept eating the marshmallows before we could recruit the participants.
Nudge blog has a description of the famous experiments along with some video – link here
In the responses to my post on why we need market-makers with big balance sheets, one thing I’ve noticed is that a lot of people are completely unable to distinguish between the argument that we need big banks, and the argument that we need the big banks that exist today. I made the former argument, not the latter. I thought this distinction was obvious, but it was apparently lost on quite a few people, who immediately pointed out that Citigroup is a very big bank, and it’s been a disaster — as if the fact that Citi was a failure somehow proves that market-makers don’t need big balance sheets.
In my post, I noted that one of the benefits of having large market-makers is that it allows the use of mark-to-market accounting, which is an important check on management. Both Felix Salmon and Ken Houghton rushed to point out that Citi doesn’t mark all its assets to market. Uh, yes, and that proves what, exactly? Citi doesn’t mark all its assets to market because it’s not required to. But that’s an issue of accounting rules, and has absolutely nothing to do with the bank size issue.
One of the things I’ve been trying to do recently is spur people to get beyond this kind of superficial sound-bite analysis. Being able to distinguish between an argument for big banks and an argument for the Wall Street banks that happen to exist today is a prerequisite for getting beyond superficial sound-bite analysis. So when I read a post like Felix’s, I honestly despair. The number of clearly fallacious arguments he treats as establish fact is just staggering, and slightly depressing. It is, ironically, a good example of what Steven Pinker just coined the Igon Value Problem: “when a writer's education on a topic consists in interviewing an expert, he is apt to offer generalizations that are banal, obtuse or flat wrong.”
On the other hand, when I read a post like this from Steve Randy Waldman, I’m greatly enouraged.
