How Goldman Sachs protects itself from a hundred-year storm

By Felix Salmon
October 29, 2012
Rob Cox today, "there’s no such thing as too much preparation."

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

(Picture from Stephen Foley)

“When it comes to natural disasters,” says Rob Cox today, “there’s no such thing as too much preparation.” He then goes on to extend the analogy:

In advance of Sandy’s march through Manhattan, thousands of sandbags have been stacked in front of the downtown headquarters of Goldman Sachs. It is a picture whose metaphorical value should not be lost on regulators, policymakers, shareholders and the bankers themselves: when the flood comes, there can never be too many sandbags, or capital, to prevent a wipeout.

There are a few problems with this line of argument. Firstly, of course there is such a thing as too much preparation when it comes to natural disasters. Cox praises New York mayor Mike Bloomberg for evacuating Zone A — the lowest parts of the city which are most susceptible to storm surges. But an evacuation of all of Long Island, for instance, or all of Manhattan, would surely be way too much.

At the same time, three’s a good reason why Goldman Sachs needed to get in thousands of sandbags: it’s in that very-high-risk Zone A.

A brief history of Manhattan skyscrapers: they were first built in Lower Manhattan, at the highest possible points around there. Look at the Woolworth Building, say, or the New York Stock Exchange, or the Bank of New York building, or City Hall, or even Chase Manhattan Plaza: all of them are on or near Broadway, which runs up the highest part of Lower Manhattan, which means that all of them are in Zone C. And as skyscraper construction moved north in the 1930s, the same thing held true: the Empire State Building, the Chrysler Building, and all the midtown skyscrapers are all well outside the reach of any storm surge.

But then skyscraper building became more high-tech and scientific, and very tall buildings started to be constructed in Zone A, very close to the water. The architects did lots of clever mathematics, or the actuaries did lots of clever sums, and soon there were dozens of huge buildings in the Manhattan flood zone; the Goldman Sachs headquarters at 200 West Street is merely the most recent.

Now, with a Frankenstorm approaching, the decision to build so close to the water is coming home to roost, and firms like Goldman Sachs are scrambling to try to protect themselves. Hence the sandbags. Which aren’t really preparation; they’re more like a desperate last-ditch attempt to save a multi-billion-dollar headquarters building from very nasty flooding.

And the fact is that Goldman’s sandbags, along with all the other sandbags being deployed up and down the east coast (including in my very own apartment building), are very unlikely to be any use at all. They’re meant to be trying to protect property against a huge storm surge, which could reach 11 feet; the chances have to be very slim indeed that the surge will be big and powerful enough to reach the sandbags, but small and weak enough that the sandbags will suffice to keep it at bay.

Or, to put it another way: when big tail events happen, the old models get broken, and you can’t rely on them any more. That’s true when it comes to building skyscrapers, and it’s also true when it comes to financial crises. In fact, it’s even more true when it comes to financial crises.

Hurricane Sandy is a known unknown: it’s approaching New York, and the only real question is how high the storm surge is going to get. It could be six feet, it could be nine feet, it could be 12 feet. Bank capital, by contrast, is something which disappears in a much less linear fashion. A bank’s capital is just the difference between two huge numbers: its assets, and its liabilities. Its liabilities are fixed; its assets are loans, and derivatives, and other financial instruments which can fluctuate in value dramatically, especially in a crisis. What’s more, assets which banks think of as being ultra-safe — “quadruple-A”-rated super-senior CDO tranches, for instance — turn out to be precisely the assets which implode in value when a crisis comes along, turning banks insolvent overnight.

And that’s just the solvency problem: the bigger issue is liquidity, in a world where banks roll over billions of dollars of debt every day. You can protect yourself as much as you like, but if your lenders for whatever reason stop rolling over your debts, you’re toast. Let’s say you needed to sell lots of US stock today, for instance. Well, thanks to Sandy, you can’t: the stock market is closed. When liquidity dries up, everybody, no matter how prepared they are, is affected, and either central banks manage to step in to save the day, or they don’t. No mere mortal, without a printing press, can hold out.

Financial crises are similar to storms: they require humility, not hubris. Being prepared can be helpful at the margin, but ultimately it doesn’t matter how good your liquidity management teams and risk ledgers and counterparty hedging operations are: if everybody else is blown over by forces beyond their control, then you will be too.

That’s why skyscrapers always used to be built well above the water level, and that’s why we used to have dumb regulations like Glass-Steagal and Basel I, which weren’t very sophisticated, but which generally did the trick. Buildings like 200 West are a bit like Basel III: they’re built with models, so that they can withstand certain forces. But if an unprecedented storm arises, they’re still more at risk than, say, Trinity Church, built more than 150 years earlier. Sometimes, simple common sense (high ground is safer, huge books of complex derivatives can blow up in unpredictable ways) does a lot more good than any amount of sophisticated preparation.


Comments are closed.