Fat Tails and Black Swans
When Morello was put away in 1910, the master of East Harlem became Giosuè Gallucci, “the King of Little Italy”. He ran the lottery, ice and coal, lent at interest and was friendly with the politicians at City Hall. He made so many enemies that, as Dash writes, ten of his bodyguards died guarding him. He himself survived half a dozen attempts on his life and was twice badly wounded by bullets.
Picture it: 1914, and you’re trying to work out whether Gallucci will be killed. The statistics are splendid: six attempts, six misses. The conclusion suggests itself. The man is untouchable.
On 17 May 1915 Gallucci and his son were sitting in a coffee shop on 109th Street that he had bought for the boy. Strangers walked in and opened fire. The son threw himself in front of his father. Both died.
Six misses said nothing about the seventh shot. All they said was that, so far, people had been shooting badly.
The Spread
We’ve talked about the average. Now for what lies around it.
Lefty the bookie did well on average. But one week he took in fifty-two thousand in bets and paid out seventeen and a half, and another week he lost thirteen. That’s the spread. Scientists call it variance. The average shows you where you’ll end up if you play many times. The spread shows how hard you’ll be thrown about on the way. Get thrown too hard, and you won’t live to see the average.
Hence the first rule of risk. Look not only at how much you’ll make on average, but at how much you can lose on a bad day. And keep your reserve for the bad day, not the average one.
When Everything Falls Together
The second rule matters more. In quiet times risks look independent. Trouble comes, and it turns out they’re connected.
Saviano writes about the wives of Neapolitan camorristi. Each of them knows her husband may go to prison, and counts on the clan to pay the family a monthly allowance, cover the lawyers, give the children something to live on. But just in case, Saviano writes, they take jobs as carers and spend hours waiting in prison visiting rooms—so as to have the means to pay a lawyer and feed the children if the clan goes under and stops the monthly payments.
When does a clan stop paying? When it gets smashed. That is, when the husbands are being arrested by the dozen. Insurance against arrest, issued by the very organisation that is being arrested, vanishes at the exact moment you need it. The camorristi’s wives understood more about correlation than some bankers.
In 1907 a financial panic swept America. Dash writes that in Little Italy alone twenty-five banks went under and twelve thousand clients lost all their savings. Twenty-five small banks stood on the same streets, served the same people and put money into the same land. In quiet times each could have failed for its own reasons. In the panic they all failed together. And for one reason.
The legitimate world learned this at Lloyd’s of London. Learned it at a high price. For centuries its underwriters had backed their policies with everything they owned, and grew rich. In 1989–1991 Lloyd’s lost more than ever before in its history—over two billion pounds a year. It lost mainly because claims for asbestos-related diseases came pouring in under old policies written decades earlier. The claims arrived all at once. More than one and a half thousand underwriters were declared bankrupt.
The Tails
And the third rule, the nastiest one. The rare thing happens far more often than the pretty formulas promise.
Scientists draw risk as a bell: most days go by normally, bad ones come along rarely, and very bad ones so rarely you needn’t think about them. The trouble is that in finance, in war and in the Mafia, the bell has fat tails. Very bad days come far more often than the bell allows.
On 19 October 1987 the American market fell twenty-two and a half per cent in a single day. If you believe the bell formulas, a day like that could practically never happen.
Eleven years later Long-Term Capital Management, a fund with two Nobel laureates in economics on its board, was doing beautifully: twenty, forty-three and forty-one per cent a year in its first three years. The fund’s models assumed that the gap between the prices of similar bonds was bound to close. At the start of 1998 the fund held nearly five billion dollars of its own and had borrowed more than a hundred and twenty billion. Then Russia defaulted. The gap didn’t close; it widened. In under four months the fund lost more than four and a half billion. The Federal Reserve Bank of New York had to bring fourteen banks together so that its collapse wouldn’t bring down the market.
Two Nobel laureates. And the model had never heard of fat tails.
The Sentence Killer
We have our own story about a model that didn’t know about the tail.
At the end of the eighties Cosa Nostra firmly believed that the verdicts of the maxi-trial in Palermo, where hundreds of mafiosi had been convicted, would be quashed in the end. Dickie writes that a view had taken root inside the organisation: Corrado Carnevale, a judge of the Court of Cassation whom the papers had nicknamed “the Sentence Killer” because of his habit of overturning verdicts on technicalities, was a guarantee that in the end everything would be all right. Past experience bore this out: he did overturn them. For years Totò Riina promised his men it would all be sorted out in Rome.
On 30 January 1992 the Court of Cassation upheld the verdicts. The presiding judge was not Carnevale but Arnaldo Valente. This wasn’t a punishment aimed at one judge. It was simply that the president of the Cassation, not without the influence of Falcone, who was then working at the Ministry of Justice, decided that trials like these would be heard in rotation rather than handed by habit to the same judge every time. The habit everything had been staked on ended with a single line in the court’s schedule. Riina lost face, and Dickie traces everything that followed back to this, right up to the Mafia’s war on the state. If your hedge rests on one man and one habit, it isn’t a hedge. It’s a bet that kept winning until it lost.
How Much to Keep in Reserve
Remember how Don Vincenzo kept a stock of oil for exactly three weeks? It wasn’t plucked out of the air. The longest any delivery had been held up, in his memory, was three weeks. I once said: so you’re covered against any delay.
The Don shook his head. “I’m covered against any delay I’ve seen,” he said. “Against the one I haven’t seen, something else covers me.” And he showed me an envelope in his desk drawer. Every year he put money into it for the day there would be no oil at all and he’d have to buy from strangers at any price so as not to let the restaurants down. As long as he lived, the envelope was never opened. The Don considered that the best proof it had been needed.
The Bottom Line
Trouble comes, and risks that looked independent turn out to be connected. So whoever insures you mustn’t be sitting in the same boat as you.