I may not get around to writing a detailed analysis of the third part, which focuses on possible regulatory measures to lessen the chance of catastrophic Flash Crash type events in the future. But the ideas raised in this part are fairly standard -- a speed limit on trading, rules which would force market makers to participate even in volatile times (as was formerly the case for market makers) and so on. I think the most interesting part by far is the analysis of the recent increase in the frequency of abrupt market jumps (fat-tail events) over very short times, and of the risks facing market makers and how they respond as volatility increases. I think this should all help to frame the debate over HFT -- which seems extremely volatile itself -- in somewhat more scientific terms.
I also suggest that anyone who finds any of this interesting should go to the Bank of England website and read some of Andrew Haldane's other speeches. Every one is brilliant and highly illuminating.
Why do we want to reduce flash crash?
ReplyDeleteIt is a huge profit opportunity for those with a time horizon longer than a few minute (i.e. buy low). Also, those with burnt fingers will eventually learn.
> why reduce flash crash?
ReplyDeleteMainly because the phenomenon is a result of market maker interactions which were intended to be a supporting and stabilizing force, not a source of risk.
If everyone could throw their own 200 lines of carefully chosen autonomous trading code into the ring, then perhaps it would be a fair game.
You are making an assumption that given the positive and negative reinforcement mechanisms I've mentioned, the system will still not adopt and self correct.
ReplyDeleteI for one would welcome such outcome since my time horizon is way longer than a few minute. I'd welcome a market that doesn't self correct!
On the other hand, I'd not bet on it. Highly profitable opportunity are very infrequent events.