Quote from makloda:
Your list above makes it look like this year's equity L/S meltdown had a historic parallel. I do not agree. "Those" panics did not have the same effects on a tick by tick basis on all asset classes as the 2007 one had. Every respective crisis affects each fundamentally different strategy in a different way. If one strategy works well in Panic A it holds no deterministic value to forecast how it will perform during panic B. E.g. Long/Short equity wasn't affected as much in 2001/2002 or 2004 as it was affected in 2007.
I am sure guys like Renaissance and Tykhe backtested their long/short equity strategies with perfectly correct EOD data (probably even with intraday data if available) all the way back into the 70s. What they could not backtest (and this is IMO that leaves them looking like idiots now) was what happens to liquidity when billions of leveraged dollars in the exact same stocks of multi-factor-arbitrage long/short equity positions receive margin calls and get liquidated. Just because a simultaneous long/short equity panic unwinding hasn't happened before (Not in 1998, in 2001, in 2004 etc. etc.) doesn't mean it is statistically "improbable" or a 25 sigma STDEV move. That's complete BS because events in financial markets are not normally distributed, plain and simple. All these multifactor quants should have stress-tested (beyond backtesting) their systems before leveraging up their investor's money five times.
Great textbook analysis...
But world class traders operate well beyond the textbooks.
All the events I mentioned were different...
But each created a general, irrational market panic...
And fearful markets have certain common characteristics...
Such as liquidity crunches and their effects.
** Many of the same historical correlations breakdown in each crisis. **
Experienced traders do not necessarily become better at making money in good times...
But they become DRAMATICALLY better at preserving capital in a difficult market.
How is this done primarily?
"Know when to hold 'em, know when to fold 'em, know when to run."
In other words... professional judgment and decisions...
As opposed to locking into a suicidal quant strategy.
Your brilliant, canned quant strategies may work well 90% of the time...
But there are obvious times when they break down...
Times that can break you.
A lot of things have NOT "reverted to mean" since last summer...
And people still waiting for "reversion" in October...
Do not have enough experience...
And are not qualified to be managing large funds.
In my experience...
At 5 years trading you get hurt in these crunches...
At 10 years you should be breaking even...
At 15 years you should be EXPLOITING the ** incredibly inefficient ** markets that result from panic .
If not...
Then the trader is are too dogmatic...
And got stuck way down on the learning curve.
Also...
There is a huge problem with "spurious correlations" in the quant world.
You can find fairly dissimilar securities that may have a high correlation for many years...
But unless the correlation is logical and explainable and expected...
And the securities are of the same industry and type and terms...
The correlation will often break down at ** the worst possible time **.
Well thought out long/short baskets are a good partial solution.
World class traders are great at avoiding "spurious correlations"...
Whereas academics and pure quants and novices and hedge fund egomaniacs...
Will always fall into this trap.
Yawn 1,000,000 times.
All the blowups make the same mistakes...
It's just not news anymore.
