It's very well chosen.
This is a very good post, and I'm glad you liked it.
A: There is no direct relationship between the maximum funding capacity of the strategy and the data of the retrospective return, volatility, etc. of the quantified strategy. How to determine the maximum funding capacity of the strategy is a matter of practical application, in general, it is difficult to say what a specific accurate number is.
Factors that may be taken into account when considering the maximum capital capacity of a strategy are: (1) the logic of the strategy itself; (2) the frequency of trading of the strategy; (3) the total daily trading volume and total holdings of the strategy trading varieties; (4) the instantaneous hanging volume of the strategy trading varieties; (5) the risk preferences of the funds; (6) the capital capacity of the same type of strategy in the market, etc.
Today is a blank and then the previous one continues.
A complete strategy for trading can be broken down into three steps:
1 building entrance
2 to hold
3 draws
The longer the duration of these two processes, the greater the total volume of transactions of the trading varieties, the less the shock cost of the strategy at the time of the transaction, and the greater the capacity.
Therefore, high-frequency trading capacity < futures trend strategy capacity < stock strategy capacity. Especially at this level of high-frequency trading, it is necessary to observe the momentary volume of pending orders to determine how many trades can be executed, and further determine how much capacity.
In this process, profits and losses arise on the balance sheet due to price fluctuations. This leads to the second main factor, the risk preference of the funds. Generally, the larger the funds, the greater the aversion to risk.
So the trend-class strategy capacity of the futures is < the cross-variety strategy capacity of the futures is < the alpha stock strategy.
However, the volume of transactions is more important in determining the capital cap, so although the risk of the futures strategy is lower than that of the pure stock strategy, the capital capacity of the pure stock strategy is greater.
Types of strategies common in the market and their funding capacity:
High-frequency trading of 1 million tons 10 million tons
Futures trend strategy of one million tens of millions of tonnes
The strategy of the futures range is 10 million tons, hundreds of billions.
Pure stock strategy tens of billions of tons
Alpha stock strategy tens of billions of dollars
This is the first time I have seen this video.
The answer upstairs basically covered what I knew, I just sorted it out and rephrased it.
Logic + data: Estimating the strategy capacity requires both logic and data. The logical perspective is to look at the type of strategy, for example, quantitative stock portfolio strategies based on a multi-factor model generally have a larger capacity, at least 5 billion yuan; whereas stock futures high-frequency trading strategies usually have a capacity of tens of millions, and the return begins to fall when more funds are added. The data perspective is to demonstrate tests, such as building a shock cost model using historical trading data.
Why the strategy has capacity on-line: The shock cost is generally non-linear as the capital scale grows, while the strategic return is linear as the capital scale grows, so when the capital scale is large enough, the shock cost must eat up all the returns, so the strategy has a capacity cap.
I'm not sure what you mean, but I'm not sure what you mean.
First of all, to solve the problem of the authenticity of your backtest data, most of the data is over-configured, resulting in self-deception.
The problem of capital capacity is generally a problem of liquidity and shock costs, please take a statistical look at the transaction data, estimate how much additional capital will increase the transaction costs, when increasing to what cost you feel is the limit, which is the maximum capacity of capital.
3⁄4 No matter how you calculate, you end up using the real disk to verify and analyze the real disk data.
This is the first time I have seen this kind of thing.
The capacity of the strategy does not need to be obtained by a drawdown, it is mainly based on the logic of your strategy.
I know the user's answer ((3 votes)
Look at the volume of trades at the time of the transaction, for example, if you trade 10,000 shares at 12:0:0 seconds, and more than 100,000 shares are traded at the same time in the market, you can say that you have no influence on the market.
This is the first time I have seen this video.
Thank you. Some thoughts.
The market size of the strategy, the volume/size of the transaction, the accuracy of historical market data, the type of strategy, the method of reporting, etc. affect the volume of funds.
2, Strategic historical testing must take into account the probability of a real-disc impact slippoint or a pending transaction, otherwise the test results obtained will be somewhat self-deceiving.
3, When I first started, I was a forward who told me the approximate capital capacity of the strategy based on experience. Over the years, I have formed my own judgement criteria by comparing the real disk with the test.
You don't know the user's answer (no votes)
A few of the top respondents basically agreed that the capital capacity of the strategy is not dependent on the return on investment, these are mainly strategic ideas: for example, hedging or unilateralism, short-term or long-term, trends or short-term fluctuations, etc.
The answer from Yang Yuan ((0 votes) is:
Thank you very much. Because I rarely know, I didn't see your invitation. The answers of the teachers upstairs were wonderful and basically covered all my knowledge of the field.
First, your question is the maximum capital capacity of the calculation strategy, first of all, the measurement of the data angle, is it fit data, or real data.
2, the logical point of view is that you build a cost model, unilateral or hedging, short-term, long-term, etc. You can build a transaction model yourself.
This is the first time I have ever seen a video of a woman being raped by a man.
The above comparison is complete, let's talk about how we felt when we did the quantification ourselves.
In fact, there is a correlation between the capital capacity of a quantified strategy and the number of shares in the strategy. If the strategy is too concentrated, the shock cost is greater when the amount of capital goes up; if the strategy is slightly more diversified, the shock cost on each share can be effectively reduced, thereby increasing returns.
Of course, if the strategy is too decentralized, returns are a big problem.
In summary, a reasonable number of shares is a prerequisite for a quantitative strategy.
The answer from Lillian Song (0 votes) is:
In general, it is two-fold.
One is to look at the volume at the time, which is generally considered to have no impact on the market if it is within 1/10th.
The second is to estimate the slippage cost based on tick data.
Then include it in your backtesting. However, in general, a simpler way is to design a function for the amount of funds and slippage, which makes it easier to design backtesting.
This is the first time I've seen this post.
I do stock quantification, generally very simple, which is to look at the portfolio to ensure that the vast majority of stocks will not exceed 10% of the trading volume on the day. Roughly speaking, this is capacity.
Translated from the original German.