No disrespect intended since I am a big fan of Quantopian’s back testing platform but I have to say the name is MOST unfortunate for such an event. There are very few actual fools on the Quantopian forum (with one or two notable exceptions) but my belief is that the vast majority of users and of people who flock to these conferences are doomed to disappointment and failure.
There are some very smart people on Q’s staff and I have learnt a lot from some of the posts and coding put up by Thomas Wiecki and other bright and enthusiastic members of the team. I have also enjoyed looking through the impossibly complex coding of Zipline. All in all Quantopian has been a thoroughly beneficial experience and I take my hat off to those guys.
I spent an hour or so this morning trailing through “Linked-Out” and my list of trading blogs on Netvibes. Many, many of those blogs have now closed their shutters and I deleted them from my dashboard. Linked-Out was full of the usual useless hucksters and salesmen pushing their equally unfortunate products. All in all a most dispiriting experience.
Except for the few commentators who add real value to my though process. Such as the excellent but not very prolific The Mathematical Investor.
But here is the real nub of the rant. The vast majority of participants in the frenzied trading world are pissing their efforts down the drain. To coin a polite phrase. If it’s outsized gains you are after, the only way to find it in probabilistic trading is to take outsized risk. And it will likely go wrong if not this week then next year, or next decade.
Outsized gains with lower risk only come where you have some sort of bias in your favour. Not a probabilistic bias based on back testing. A real bias: the bid offer spread, arbitrage of some sort, buying dirt cheap distressed debt, private equity and so on down to the naughty world of front running and insider dealing.
The vast majority at such conferences (be it the unfortunately named QuantCon or any other) can at best hope to receive expert help in coding, statistics or machine learning. And that is at a good conference – I imagine QuantCon was good; or at last far better than the majority.
But when it comes down to it, unless you can find a bias in your favour you are wasting your time. You can’t predict the future and most if not all “systems” will crash and burn.
Hence my emphasis in recent years has been on simple asset allocation and simple re-balancing strategies. Economies are still growing and so are corporate profits. Real estate is an asset with a fixed and diminishingly available supply. Bonds are essential to even out volatility. Commodities may not offer much actual growth but you might benefit from contango or backwardation and the effect of diversification.
You are not going to go too far wrong with a portfolio constructed from a few liquid physical ETFs covering world markets and asset classes.
If you are a long term investor any other route is a farcical and expensive waste of time. If you are a hot shot trader forget about probability and stupid ephemeral patterns and look for a real market niche with real profits.
No offence intended – especially to the bright boys at Q. I am merely working out and confirming my own thought processes and views. With which, I have no doubt, some will find violent exception.