The King of Neutral Strategies - Analysis of Market Making Strategy Logic (Part 2)

To solve the problem of breaking the grid, we will implement a simple hedging strategy:

When we buy 10 apples at a price of 10 yuan, we also short 50 apples at a price of 10 yuan.

Let's go back to the previous situation:

The price of apples fell from 10 yuan to 5 yuan. At this point, we have 55 apples and 465 yuan in cash. However, we shorted 50 apples and realized a profit of 250 yuan.

Let's calculate the total assets:

55 * 5 yuan / piece + 465 yuan + 250 yuan = 990 yuan

Hey? No, that's not right. Compared to our initial total capital of 1000, we are still in a loss:

(990-1000)÷1000*100%=-1%

This is not a losing strategy, is it?

(Thinking question, how is this 1% loss caused? What parameters does it depend on?)

But don't forget, our grid strategy has been trading all along.

Assuming the price fluctuates around 10 yuan, as long as there are 20 trades (10 buys and 10 sells), we can realize a profit of 10 yuan, which just covers our previous risk exposure.
From the 21st transaction onwards, any fluctuation that occurs and leads to profit for us is considered as pure profit.

We turn this vague oral statement into a precise mathematical definition:

20 trades, the average price is simply calculated as 10 yuan, then the volume is:
20 times x10 yuan/time = 200 yuan.

I will introduce one more parameter, called Turnover Rate:

Volume 200 yuan/Total capital 1000 yuan = 0.2

In other words, in this strategy, if the turnover rate exceeds 0.2, the strategy can achieve fluctuation profit to offset the exposure of hedging (i.e. 1% loss).
In the example of my previous experience as an MM (Market Maker) at an exchange, within less than a month, we generated a trading volume of $324 million with assets of $249,000 USD, which accounted for 36% of the initial $690,000. The turnover rate was approximately 43 times per day on average (based on data from the past month).
I believe the explanation of this principle should not be difficult, you should understand it. The principle is indeed simple, and making profits is not difficult.

Don't rush, let me give you an example here:

I have never been in contact with the fren of Texas, and I can play with us all night, after all, there are 2 cards in hand, and we are generally small cards below the gourd, and we will know it by playing two games, which is very easy to get started, and there is no difficulty.

But every player who is familiar with it, and has watched a real poker game, knows that, seemingly simple Texas Hold'em, the calculation of cards and probabilities is far beyond the imagination of ordinary people.

Similarly, seemingly simple market-making strategies follow the same logic. Let's consider these issues layer by layer.

"First layer:"
How to create a rigorous mathematical model for the description I provided of the model of apples and cash?
How to establish a model relationship between the grid interval and the amount of funds used?
This mathematical modeling is actually the simplest, but I believe it can stump 90% of people.

Second Layer:
How to control trading slippage?
How to negotiate transaction fees and liquidity service fees with the exchange?
Does the trading program have enough robustness to handle extreme market conditions, even in the event of exchange downtime?
This layer mainly aims to control external factors. Without a certain amount of funds, there are no conditions to completely eliminate the impact of these external factors.
Whether the strategy can be profitable smoothly, these external factors have a huge impact, and it is very likely that the deviation between theoretical profit and actual profit is mainly affected by these external factors.

Layer 3:
How to share funds and trade the same funds in more than 20 different varieties?
How to set the parameters of volatility for historical data and then let the program automatically calculate the grid parameters and adjust them?
How to set the fluctuation range? How to find the optimal solution between maximizing the use of funds to earn high yields and risk exposure, as well as the impact of factors such as fees, slippage on the parameters?
In this layer, because multiple parameters are intertwined together, it has become a chaotic model to some extent. The computational complexity is enormous, and it can only be addressed through complex mathematical modeling or AI.

However, in any case, along this line of thinking, at least the development of quantitative strategies has moved away from the guessing game and entered the right track of developing quantitative strategies in a scientific and rigorous manner.

Hope this strategy sharing can help you.