Quantitative trading is not guaranteed to earn and there is nothing too tall! Demystify quantitative trading

Quantitative trading is a popular term in the field of financial trading in recent years. The so-called quantification is the exponential quantification. Quantitative trading is to provide traders with trading basis for trading behavior in a quantitative form, so that the trading results can be excluded and avoided as much as possible from the arbitrariness and psychological fluctuations of subjective transactions.

Quantitative trading has been practiced in the United States for more than 30 years, most notably the medallion fund of mathematician Simmons and his Renaissance company, which has averaged an annual return of 35% since 1989, and the medallion fund has been hailed as the most successful hedge fund.

Quantitative trading in a narrow sense is a programmed transaction that has begun more than ten years ago, it is the trading method used in the trading process, using computer language to compile computer software programs, to achieve machine stock selection, automatic order trading and other behaviors. Through the computer program, some labor costs can be saved (slow manpower analysis, after all, there are more than 4,000 stocks in the market now, and there will be more in the future), and at the same time, some traders can avoid unnecessary time to keep an eye on the market, and also avoid the influence of emotional and psychological factors to a certain extent.

In a broad sense, quantitative trading is the systematic trading that we traders use in the trading process. A systematic approach to trading according to some fixed trading models, systematic trading is a prerequisite for stock trading profits. For example, the fundamental value investment method quantitatively sorts out many financial data and indicators into a fixed model, which belongs to fundamental quantification; people include various indicators compiled by using technical analysis theory, stock selection conditions, etc., which belong to technical quantification;

In addition, quantitative trading is divided into algorithmic trading (that is, high-frequency trading, mainly used for grabbing orders), arbitrage trading (inter-period arbitrage and cross-variety arbitrage of futures varieties), computer programs that realize fully automatic trading compiled according to the existing various technical analysis theories, and so on.

Quantitative trading is not a guarantee of profitability, it must be based on a certain probability of success model in order to be applied in practical trading. We all know that the root of the casino profit is actually 1% higher than the probability of the player's profit, which is 1% higher than the probability of profit to ensure that the casino will win for a long time. So quantitative trading is actually pursuing a 1% higher probability than most people in the market. But this 1% is not something that the average investor can do. It takes a lot of practical summary and review summary to finally form the so-called quantitative trading model.

Finally, investors are reminded that the quantitative trading model mainly comes from the following two models:

1, data mining, from the historical data to find a large probability of profit in the past history of the model, this model is generally a black box model, black box is that you can only see the results, do not know the logic, such as the popular machine learning model, is a typical black box model. Its disadvantage is very obvious, that is, you do not know the principle of profitability, whether the probability of continuing to appear in the future in line with the above model is, that is, this model, the historical performance is very good, but whether it can be profitable in the future is very uncertain.

2. Derived from the subjective trader's profit model, according to the systematic trading method of the subjective trader who is profitable, the trading program is compiled in computer language. This trading model can be quantified, some can not be quantified, if the part that can be quantified is more, and the probability of profitability of the historical data after quantization is high, then there is a high probability that it can be used for real disk. Unfortunately, this model is rare and cannot be sought. The other is that a small number of parts can be quantified, most cannot be quantified, and the energetic part performs poorly in historical backtesting, and most of the profits of subjective traders may come from subjective judgments, which accounts for the vast majority of this model. For example, Xu Xiang's up-and-down board death squad board model, buying on the stop-and-go board can be quantified, but if it is only up-and-down board buying, it can not achieve profits, and the greater reason for profit lies in the so-called disk sense of the trader, so the mining of these disk senses is the key to this type of model.

In summary, quantitative trading is only a small branch of trading, not a profitable method. Don't be superstitious about so-called quantitative trading.

In fact, the systematization of transactions is the key, and the key to systematization is to try to objectify the quantity of subjective transactions as much as possible. Good luck with your investment!