At least you can recognize the first point: the lack of black swans can be a black swan itself. And I don't necessarily disagree with your methodology or your reasoning behind it. But as Taleb would say, your personal results are nowhere near statistically significant. So even if you are 'intellectually sound' in your theories, a empirical skeptic would have to wait for the results that you provide before determining whether or not you truly have an edge. You know where I think Taleb goes wrong? When has their even been a black swan that has been POSITIVE for the market. When does the market suddenly jump up 20% in a day? Never. But it sure has gone DOWN by that much. Black-swans have always been psychological effects driven by FEAR. His methodology is to basically buy options that he claims are mis-priced because people call them high-sigma events, when fat-tails are actually more common than we claim. However, the fat-tail to the negative side is far larger than the positive. So why not just WAIT until markets start trending down before buying the insurance options? You would save a whole lot of premium. Hell, even October 1987, which was considered a very 'RANDOM' event, was pre-cursed by a very pronounced down-turn. I find Taleb to be sort of funny. He writes off all other managers and their results, claiming they have been fooled by randomness, but he has yet to generate a statistically significant record himself. Furthermore, the ideas that Taleb has presented are no more 'statistically significant' than many other ideas. Can I not justify his presences as a factor of statistics, just as he justifies Buffett's existence? If every year, half the theorists were dropped from our pool because their theory no longer fit the model, eventually we would be left with a few theories that fit really well. But that doesn't exclude the fact that a black-swan could come along and disprove these theories in one swift blow. As you said, the black swan of NO black swans ever coming again would prove Taleb both right and very wrong. Taleb's OWN ideas should preclude him from claiming any sort of significance in his own theories. How ironic. So I am not trying to say that you are in anyway wrong in the method you are taking. I just find all of this highly ironic. Fundamental theory does not supersede statistical significance. Nor does statistical significance supersede fundamental theory. They MUST go hand in hand. "It works in theory, but does it work in practice" must be married to "it works in practice, but does it work in theory." Does your methodology satisfy both? Sorry for the thread jack. Carry on.
Good post. Actually I don't think it's a thread jack as system trading is system trading whether you are trying to catch 1 year or 100ms trends. Your system has to back test well over as much past data as you can get. And crucially, to back test well, you need low variance. Back testing a great average return itself means squat. Taleb is a great writer and a mediocre fund manager. He has the right basic idea but fails on the details. In essence he is trying to be continuously long a straddle in order to profit from big moves (not necessarily market crashes) without knowing when they will occur. The problem is, he does this by buying options which cost him an arm and a leg compared to what they pay off. Hence, he can't make any money in most years. Good systems duplicate owning the options far more cheaply by trading the underlying. No positive black swans? What about the increase in the oil price from ~$45 to ~$150 in 2007/8? That's a classic example of a long, slow black swan. It doesn't have to be a one day event - that's a red herring
Or is that just a trend? I suppose the definition of a black-swan depends on your time-frame? A 20% drop in a day for a long-term trader is a black-swan. If I looked at a 20 year chart, would a 200% run-up in two years be a black-swan as well? Maybe it would...
Hmm, I think the essence of Black Swandom is how generally *unexpected* the move was. I think that oil rally really shocked most people. The market had been rangebound for years before that, prompting a slew of academic papers explaining why the oil price must remain in that range My system loves trends/slow black swans but can't necessarily do much with 1 day events. It only looks at EOD prices.
Wrong, numerous markets have gone up a lot on news. Also, Black Swans are not normal market events. You guys are talking about "Grey Swans" like rate cuts, market crashes etc. These things happen every few years and are not surprising at all, they are a normal part of market activity. A black swan would be something like the S&P crashing to 17, then rallying to 57,000 in one day, then going back to unchanged on the bell. I could be wrong, but I would think it's possible that no financial market has ever had a black swan event.
No, the point of a black swan is that most people don't think they exist, until they do. Taleb brings them up because just because you haven't seen one doesn't mean you won't. Just like seeing a chicken alive for 1000 days in a row doesn't mean the farmer won't cut its head off tomorrow. Specifically, Taleb points out that these '10-sigma' events are not really '10-sigma' events because markets do not follow normal distributions. Instead, they have fat tails. So 10-sigma really becomes 4 or 5 sigma, which totally changes how we should be approaching risk. These aren't grey swans. They are low probability occurrences that happen far more frequently than we give them credit.
lolatency, What you are trying to do is very difficult and will only work for a few people .... but it is possible. Just keep your head down, do sensible things, avoid blow ups and keep your cool even as your pnl swings up and down.
Precisely. Stock market crashes happen once every few years but each time one occurs the newspapers holler that it's the end of the world as we know it. That's because stocks trending upward is what everyone thinks is "normal".
swans back or forth, I dont see any edge whatsoever in being on balance long vega over prolonged periods of time, and I say this as a prop options trader. In an ironic sense I almost would say you have an edge if you can afford yourself the luxury to be a consistant option writer. In that sense you are in fact long the call. You collect your annual bonuses on the market making desk, every now and then you completely blow up, get fired, and start your new job a few months later after having taken a nice vacation and time out. Thats pretty much the approach a lot of sell-side options traders take. If you happen to be the most liked guy in the bank you may even get away with grounding 10-20 million. Back to the topic, would anybody share, without giving details away, where they see an edge in high frequency trading without access to order flow? I still fail to see where those who claim to be able to run high frequency on the retail side generate alpha. Thanks