Tuesday, December 23. 2008
Bernie's histogram
Madoff vs GATEX monthly return histograms and density estimates, 1993 - 2008
The Gateway Fund has been using Bernie Madoff's collared beta strategy for almost as long as he supposedly did. GATEX's monthly returns indeed look like an investing time series: fat-tailed and narrow-peaked compared to a normal curve. Madoff's don't. Madoff's look just like a normal curve, except that if you squint, they are slightly positively skewed, and their negative tail is too small. I would love to invest with someone who actually could produce returns like that, but collared beta doesn't.
Does Jim Simons?
Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
Madoff's strategy was never split strike conversions. The structure was an unsecured loan to a trading co. and the split strikes were used to pass back an interest rate (a form of finite reinsurance if you will). The split strike was fabricated to generate a specific rate of return. I say fabricated because many times the prices were not simultaneously available in the market.
(I did due diligence on Madoff in 2001 - it was clear to us that we couldn't see the collateral and didn't understand the underlying process that generated the returns, so we recommended against it.)
Don't feel bad, most people make the same mistake in thinking the split strike conversions were the strategy.
That's why we got paid to do the DD!
(I did due diligence on Madoff in 2001 - it was clear to us that we couldn't see the collateral and didn't understand the underlying process that generated the returns, so we recommended against it.)
Don't feel bad, most people make the same mistake in thinking the split strike conversions were the strategy.
That's why we got paid to do the DD!
Very interesting. The press had me thinking it was simple collared beta.
The structure you describe is even more suspicious than a well-understood process generating an unexpected distribution. True, 'passing back' an interest rate could certainly generate these returns, but raises the question: why structure unsecured loans in such an odd and opaque way? Why use options at all? What FASB justification is there for using nonmarket prices for liquid instruments?
And yeah, DD has its moments. Though with Madoff, the more I learn, the more it puzzles me that anyone with a few dollars to invest didn't smell a rat.
The structure you describe is even more suspicious than a well-understood process generating an unexpected distribution. True, 'passing back' an interest rate could certainly generate these returns, but raises the question: why structure unsecured loans in such an odd and opaque way? Why use options at all? What FASB justification is there for using nonmarket prices for liquid instruments?
And yeah, DD has its moments. Though with Madoff, the more I learn, the more it puzzles me that anyone with a few dollars to invest didn't smell a rat.
Well, you couldn't say for certain it was a rat, and at one time it wasn't.
The premise most professionals had for believing their might be something legit was this. "Madoff was borrowing to fund his market making ops. Those ops made good returns. Good enough that he only wanted to give away a more or less fixed amount, and pocket the rest."
If he had a 40% year, he paid 12% and if he had a 7% year, he paid 12%, making the difference out of retained earnings." I would have thought this was the case through 1997.
So if the real monthly return was say
7% 2% -4% 2% 3% -5% 1% 0% -2% -1% 2%
then he paid 1% 1% 1% 1% 1% etc.
This gave him the sharpe he needed to borrow size, but didn't reflect the underlying risk process.
I believe he got killed by double compounding. The laibility at ~12% growth, while any annual sequence of capital loss would have had a negative compounding effect on the capital base and made it virtually impossible for the smaller capital base generate returns capable of supporting the liability base. (sound familiar? think social security, or the budget deficit).
My argument has always been if his risk looked like the returns he gave investors, then why would he be willing to give it to them in the first place?
He could have borrowed in the capital markets MUCH cheaper. So, if you can't figure out why you are so lucky, you might want to consider why you are so dumb.
BTW there was not enough evidence to say it was an outright fraud, only that it didn't make sense.
Now these same investors would like someone to bail them out and give them the fictitious compounded return that they were so sure they were entitled to.
Madoff is not the only troubling element of this entire affair.
The premise most professionals had for believing their might be something legit was this. "Madoff was borrowing to fund his market making ops. Those ops made good returns. Good enough that he only wanted to give away a more or less fixed amount, and pocket the rest."
If he had a 40% year, he paid 12% and if he had a 7% year, he paid 12%, making the difference out of retained earnings." I would have thought this was the case through 1997.
So if the real monthly return was say
7% 2% -4% 2% 3% -5% 1% 0% -2% -1% 2%
then he paid 1% 1% 1% 1% 1% etc.
This gave him the sharpe he needed to borrow size, but didn't reflect the underlying risk process.
I believe he got killed by double compounding. The laibility at ~12% growth, while any annual sequence of capital loss would have had a negative compounding effect on the capital base and made it virtually impossible for the smaller capital base generate returns capable of supporting the liability base. (sound familiar? think social security, or the budget deficit).
My argument has always been if his risk looked like the returns he gave investors, then why would he be willing to give it to them in the first place?
He could have borrowed in the capital markets MUCH cheaper. So, if you can't figure out why you are so lucky, you might want to consider why you are so dumb.
BTW there was not enough evidence to say it was an outright fraud, only that it didn't make sense.
Now these same investors would like someone to bail them out and give them the fictitious compounded return that they were so sure they were entitled to.
Madoff is not the only troubling element of this entire affair.
Yeah, I'd have passed on the structure alone, much less the distribution. You want me to give your private business in a cyclical industry with high technical and operational risk an unsecured loan at 12% with some sort of coupon flexibility to boot? I had better be senior to bank debt, see your books every quarter and get some extremely cruel lawyers to draw up the paper.
Me, I don't think it ever wasn't fraud. Bernie may even have fooled himself for a while, but the more I hear, the more through-the-looking-glass I feel. I think he got killed because he set up a Ponzi scheme.
As for Social Security and the deficit, the things to remember are that societal pension schemes pay retirees out of current production, full stop, and that spending is taxes. Everything else is accounting and politics. Those are big things, but there is no Ponzi involved.
Me, I don't think it ever wasn't fraud. Bernie may even have fooled himself for a while, but the more I hear, the more through-the-looking-glass I feel. I think he got killed because he set up a Ponzi scheme.
As for Social Security and the deficit, the things to remember are that societal pension schemes pay retirees out of current production, full stop, and that spending is taxes. Everything else is accounting and politics. Those are big things, but there is no Ponzi involved.
why structure unsecured loans in such an odd and opaque way?
Because if people thought it was a loan, they would have required it to be secured by the collateral. By couching it the way it was, money was more easily obtained.
Being structured as a trading company had many advantages over as a customer (reg T, market-making rule preferences, short sale, stock loan to name a few)
Because if people thought it was a loan, they would have required it to be secured by the collateral. By couching it the way it was, money was more easily obtained.
Being structured as a trading company had many advantages over as a customer (reg T, market-making rule preferences, short sale, stock loan to name a few)
What collateral? A market maker's collateral is some rapidly depreciating computers and some software. If he blows up, the software must have been useless and the computers will be worthless tomorrow.
But yeah, I can see Bernie selling it as regulatory arbitrage. Like AIGFP that makes me itch, but there's no business more profitable than cashing those government checks while they last.
But yeah, I can see Bernie selling it as regulatory arbitrage. Like AIGFP that makes me itch, but there's no business more profitable than cashing those government checks while they last.
Quicksearch
Categories
Syndicate This Blog
Blog Administration
Powered by
