Madoff's stated strategy was split strike conversion. According to Madoff, this involved going long a basket of stocks against buying puts and shorting calls. These are very closely watched arbitrage relationships. For example, long SPY stocks, of course, tracks the SPY instrument. If it does not, virtually risk-free arbitrage quickly bangs it back into line by going long the individual stocks, and shorting the SPY instrument. And, long the SPY stocks has a similar, very tight relationship to the SPY options, because a basket of long SPY stocks, can be counterbalanced almost exactly through the creation of a short SPY position by going long SPY puts and shorting SPY calls. Even if a broadly-diversified, random basket of stocks is selected for this strategy, the returns should converge to the S&P 500 over time (think of how hard it is for the average fund manager to surpass the market averages over time - or even to significantly lag them on a risk-adjusted basis for a diversified basket of stocks over time). Long stocks, combined with long puts and short calls in a delta neutral ratio, (which is what Madoff was trying to simulate through the very consistent returns), will produce very stable returns. Just as it is unlikely to make much money with this strategy, it's also unlikely to lose much money with this strategy. A $50 billion loss is not within the realm of possibility. Consequently, Madoff was not pursuing his stated strategy. The strategy evolved into something else over time - possibly to catch up to past losses, or the money was stolen. Expect Madoff to admit he was not actually trading in the manner he described, or, more ominously, expect him to say he just "winged it" and no records were kept. The second option is an irresistible cover for theft, in the days before the fraud becomes known.