Financial markets are very wide. They include stock markets, Forex and commodity market, derivatives market, and so on. Which of them are most preferable for trading from your point of view?

At each moment of time there are markets that are most preferable for trading, but it’s not easy to detect them beforehand. For example, the lowest cost for market entering for an individual is in Forex; the Russian stock market has been the simplest and least risky one for quite a long time; for large sums of money American bonds are suitable, for advanced traders American stocks are ok (though not now), etc. And from the point of view of minimal risks to a contractor the most optimal solution is stock holdings, for example futures.

Let’s dwell on the idea of “an efficient market”. Do you think regularities exist in markets? And can one make money using the detected regularities of market price movements?

In my opinion, the classical efficient market theory, especially in its basic notions about rational investors, their equal information awareness, etc. is far from reality. As well as the conclusion that the stock mechanism (or the expanded net of over-the-counter trades) adequately reflects asset prices. But the conclusions are true in many respects, if we substitute the word “efficient” with “accidental” and admit that the market is almost always wrong, but we have no other mechanism for an adequate pricing. This is like in Churchill’s famous dictum: “Democracy is the worst form of government, except for all those other forms that have been tried from time to time”.

Nevertheless, statements that price time series are random and in extreme case appear to be martingales are also incorrect. In one of forums we managed to reach a consensus formulated the following way: prices resemble very much random walks; however, actually they are not. So, there are ways to make money: there are no laws, I think, but there are regularities, helping to earn.

If there are such regularities, can they be described mathematically or formulated in any other quite strictly determined way?

We can talk about existing regularities only if we managed to somehow formalize them and prove their static validity. Otherwise they remain our hypotheses or – in most cases – only illusions, even though they are working ones. But to find such regularities, formalize them and use suitable programs for checks are three quite complicated tasks.

What is, in your opinion, the nature of regulations origination?

Regulations do not originate from nowhere – they are always determined by the nature of things and interaction of objects in complex systems. It is recommended, though not strictly obligatory, to have an idea of what is underlying this or that regularity. Besides, the character of regularities can change from market to market – there are common features, as well as there are essential details. It’s no paradox, though it is usually asserted that if regularity exists, it works at all times and in all places.

For example, excessive money liquidity entails the growth of stock markets – one has to do something with money. And if the crisis of money liquidity happens (as, for example, in 1998 in Russia), stocks fall below the levels that could be imagined in all our nightmares. This is also a regularity that changes old regularities.

Do you think markets change in the course of time? Is there any difference between the market of today and the market of 10-20 years ago?

Markets like anything in the sublunar realm are subject to changes. Naturally, almost all present-day markets differ from markets of 10-20 years ago. But this doesn’t mean that there are no more common regularities that were, are and will be. Though, they can be diluted and show other numerical characteristics at their parametrization.

What in your opinion should be the required qualification for a trader to successfully trade in financial markets? What should such a trader study?

If a trader is willing to learn, he must first of all study. But primarily a trader must have common sense and creativity. From my point of view the formal higher education in economics does not play much role here, though general knowledge and acquaintance with mathematical apparatus are undoubtedly advantages.

You should from the very beginning understand that market is a very competitive and quite noisy area and try to detect those of your character qualities that would give you market advantages. And they should be further developed. You can also find a job on the other side of a counter – for example, a brokerage company. In such a case you can try to understand something at the expense of an employer and then if you are still willing, can try to carve your own career.

Generally, reading various (but exactly various!) books in economics and trading, analyzing competitive points of view including those in the Internet, studying mathematics, probability theory, mathematical statistics and programming language can be recommended to everyone.

I believe that human sciences are also helpful – we need to understand the market and motives that actuate its participants and regulatory bodies, and for this purpose we should understand life in the wide sense.

Do you know programming?

I know a little about programming which allows setting reasonable tasks to programmers with whom I work. As for programming, for me it is an excessive and absolutely unnecessary load.

Are these tasks connected with the automation of trading (speculative trades in the market)? What is your attitude to the automated trading?

My attitude towards automated trading is positive. However, one should understand that the primary notion here is the robust trading algorithm, and its automation is only a technical task, though not always easy-to-implement. Automated trading is mostly efficient on short timeframes where it is hard or even impossible for a human to quickly analyze or execute signals. So, one can entrust trading to robots, provided there is a robust algorithm and strict control.

Limitations here first of all are connected with limitations of the algorithm itself – of the trading system and market liquidity in the working timeframe via a selected broker. Additional problems may occur if a system is so fast that a broker has no time to move your risks to the market – no one is willing to pay out of one’s own pocket.

What do you think about optimization of parameters? Is it ‘evil’ or a mandatory condition in the process of a system creation? And where is that thin line between optimization and fitting of parameters?

Don’t confuse optimization with “OverCurveFitting” – these are two completely different things. “OverCurveFitting” usually occurs when there is no trading idea or numerous absolutely unimportant facts are added to it. This is so when we want to do without back testing on periods that are not used for optimization. In other words, a trading idea even if without a clear formulation is primary. So a reasonable optimization is necessary.

There are ways to differentiate between them: the main one as I’ve already told is the check of a basic trading idea (don’t mix up the idea with a crossing of two MAs with different periods) and real-time testing, other ways are diverse and quite complicated. What we need here is the maximal self-criticism – do not try to start working directly with a substantial sum of real money. Of course, working with good deals of money gives inestimable experience. However, try to make this experience as cheap as possible, though anyway you will have to pay for it.

In the Automated Trading Championship numerous parameters are calculated for each Participant to analyze the quality of trading. Which of them are the most important in your opinion, and what parameters could be added?

I think there is a lot of excessive information – for example, Gross Loss and Gross Profit, which are quite usual for statistics of most testers. For me personally of greater importance is the estimation of annual profitability and, what’s even more important, maximal drawdown. Besides, profit factor should also show good values. A disadvantage here is that all parameters, especially drawdowns, concern balance, not equity. I.e. maximal drawdown inside a trade, not for a closed position is the main indication of a trading quality. It means profitability and maximal drawdown for equity are the key parameters.

Sharpe, Sortino or Calmar ratios connected with them also offer a descriptive characteristic, but in this case formulas for their calculation should be indicated to avoid misinterpretation. The number of trades and profit-factor are significant. All other characteristics are useful, but mostly for a deep analysis. And, definitely the best visual assessment can be obtained from a graph of account equity applied to a time series of a basic asset prices.

Did you know anything about the Championship earlier? Do you have any wishes to its Participants or Organizers?

Yes, for two years I have been observing its development. It is an interesting project. By all appearances (and it’s not so much their rating in the competition as their answers) the winners of both Championships have fully justified and interesting results.

What I can wish is the further active development of this theme – not only Championships, but also the sphere of development of trading methods and improvement of the platform. I am not wishing luck – it is never enough to go round.

Thank you for the interview. We wish you good luck!