How to Exploit the Volatility with Algorithmic High Frequency Trading
I start with volatility because it has become synonymous with risk. The first reason is a significant boost of the academic world. The second reason is the immediate impact of volatility on our personal feelings and our desire to respond immediately. It is to our defense that all the commotion around us, the news and the interpretations, are almost always created by volatility, not by price.
Any research that deals with the capital market in one form or another must address the risk in some measure. Volatility is the easiest and most measurable thing to measure. And it soon became a standard.
There are several failures in this automatic reference that do not get enough explanation.
The most prominent of these is that the ratio to volatility is personal, changes over time and is intuitive. Each responds to fluctuations in a different way. Usually not in a calculated manner, and almost never in a predictable and predictable manner.
The second is that volatility is usually positive when it has a positive bias in relation to yield. That is, if over time a particular property rises in price, the fluctuations on the way must be more positive than negative. Which means that negative fluctuations can increase the yield over time over its average, assuming there is enough money and discipline to take advantage of them. This is not true for all instruments, but it is valid for investment portfolios that are built over time in algorithmic trading, and products (dividends, bonds).
In other words, it is problematic to present volatility as a negative or a disadvantage to be avoided.
Assume two properties with the same expected yield, one of which is 2 times more volatile than the other. See, for example, the graph below, the continuous line describes a more volatile asset than the asset described by the dotted line.
Theoretically, an investor should prefer a less volatile asset. We tend to think that the second asset is more risky, with a low return that does not compensate for its risk. In practice, the more volatile asset can give a higher yield over time, especially if the investment model is algorithmic. It differs from investor to investor and depends on the form of investment.
The third failure is at the point of assumption: each risk is expressed in volatility, and all volatility is caused by a change in asset risk (in other words, a more volatile asset is a more risky asset). These two sentences are not necessarily true. Systemic risks and rare events are almost never expressed in price. We do not have the ability to evaluate them, we do not know their frequency, we do not have the ability to know their effect in advance and they influence different properties in different ways.
Risk is created (according to Elroy Dyson) that there are more cases that can happen, than those that lied at the end. We would expect volatility to reflect this. Ie properties that have a wide range of possible outcomes will be more volatile all the time.
But that’s not the case.
If volatility would have reflected risk, a risky asset would have been volatile all the time. The list of risks and the effects on it does not change. In practice, most of the volatility is caused mainly when assets fall in price.
Take the US stock market index, the S&P500. The graph below shows the distance from the last peak in percentages, and the Y axis the average volatility appropriate for that fall (I used VIX and a peak of 100 days, but other measurements of back fluctuation and different periods gave similar results).
Volatility is growing as we get further away from the peak. The correlation between the VIX and the market is so great that to say that VIX is the measure of fear is to say that the market is the measure of fear.
Anyone who believes that volatility is equivalent to risk should also believe that the stock market is more dangerous as it is far from the peak. Obviously this is not true.
Another thing that no one can do for you is knowing how you will react in the event of a decline in your portfolio.
Investors do not intuitively think of volatility, but do react in such a way. In other words, there is no way to translate concepts like standard deviation into the feeling of fear that will come from the loss of money that day. There is also no intuitive way to compensate for volatility in yield. No investor thinks “I’m willing to be afraid to some extent for a 5% return, and I’m willing to be a bit more afraid for a 6% return, but I’m not willing to fear more than that for a 5.5% return.” This sentence sounds a bit ridiculous, but in practice it’s the way many investors, consultants, books and studies talk about risk. In other words, they create a dissonance that must then be bridged when reality slaps in the face.
This is one of the reasons I almost completely avoid measuring volatility. I prefer to use terms like “maximum reduction” or “maximum drawdown” which is easier to imagine, especially if you convert the percentages to dollars.
For example, if portfolio 35/35/30 experiences an extreme decline of 21% and the average decline from the peak is 3%, then each investor should expect some 12% decline sometime during the investment period. This is a more obvious form in my opinion to think about losses. It allows the investor to better imagine the experiences and feelings he will have to undergo during the life of the case.
An investor will work wisely if he imagines losses that are even worse than 1.5 or 2. No matter how long the history of these cases is, we never really know what is around the corner.
Coping and Exploiting the Volatility
Now we know that it is not possible to measure volatility intuitively, and we know that it is impossible to know how these losses will be seen. Therefore, it is worthwhile for the investor to strengthen the only area that he controls (a little), and this is his coping with fluctuations and temporary losses in the portfolio.
Part of the ability to cope with losses depends on the nature, and this can not be changed. At least not easily. But there is also a part that can be trained. To try to do thought exercises, to force inactivity, to avoid external influences such as newspapers, commentators, forums and hallway conversations.
The main aim of all the noise around is to convince the investor that the loss he sees in the case in the morning will only increase during the day and eventually become a permanent loss that will never be covered. The answer is that such a chance always exists, and only a fool will say it will never happen, but a balanced global portfolio has never lost. The investor has a much greater chance of catching a flight that will be hijacked, crashed or just taken.