Overview:
Yesterday I briefly discussed my spot grid strategy, and some friends said they didn't understand it. Today, I'll organize it. In this AI era, every company has its own quantitative trading, and personal thoughts should not be kept to oneself.
Currently, my dynamic spot grid strategy is running well, with stable daily income (1000U investment yields a stable income of over 5U daily, over 10% monthly income). So far, there has been no situation of capital being trapped at high positions (last night the price rose, and all trapped funds were sold out on 2025.04.17).
Three core strategies:
1. Set grid positions and grid ratios according to different price ranges. The higher the price, the less capital invested, and the larger the ratio style. This prevents a large amount of capital from being trapped at high positions.
Taking #SOL as an example
Price range 135~140, place 150U per grid, 1.0% per grid.
Price range 115~130, place 300U per grid, 0.9% per grid, as shown in the figure below.
2. Dynamic adjustment. Automatically widen the grid ratio during rapid declines. (There are many strategies, such as tracking price changes over time or simply judging whether to buy continuously without selling.)
3. There will always be only one buy limit order. There is no need to place all grid buy limit orders in advance; set the buy limit order based on the current price, the status of current orders, and dynamic grid ratios. This avoids a large amount of capital being tied up in buy limit orders.
Cryptocurrency selection:
Among the top six, I chose #SOL because its activity and stability are both acceptable. Currently, #FDUSD has no order placement fees.
Grid and capital:
The grid setup should cover at least 15%. If the chosen cryptocurrency is stable, 10% is also acceptable. Calculate the step grid costs and density based on your own capital (denser is not always better; transaction fees are very high and occupy a lot of capital).
Capital planning:
Capital planning is an important factor determining grid effectiveness. Generally, use 1/3 for the grid, 1/3 for standby at high positions, and 1/3 for bottom-fishing funds. Most of the time, only 1/3 is in use, and the remaining funds can be used for some wealth management. Therefore, a principal of 1000U may yield the same return as 3000U.
PS:
Friends who don't know programming or have only a little foundation can directly let ChatGPT generate Python code for my strategy. Be careful about price quantity precision issues in the code that may cause order failures. I won't provide the code; develop it yourself and bear the risks yourself.
I don't like DeepSeek; it's purely a personal preference. Don't come to me with which AI is better.
