TLDR If the price goes up at an ever increasing pace, increase the weight of that token in the pool
Works best with: Large cap cyclical coins and tokens
The father of this strategy once said: far more money is made buying high and selling at even higher prices than buying low and selling high. Momentum trading is a staple of quantitative hedge funds and Commodity Trading Advisors. More recently, momentum ETFs have arisen as it can be argued that the level of active management needed is minimal.
How it works
Importance of price gradients
The rate of change can be very telling of how a price might change in the future. A token that is increasing at a greater rate every day is considered to have a strong positive price momentum.
The vanilla momentum strategy increases the weight of a pool constituent if there is a strong positive momentum signal and decreases it if it has a strong negative momentum.
Momentum works best when there are strong bull and bear markets forming market cycles. As Cryptocurrencies have followed this pattern this strategy has historically performed well.
An important question is what time period should we consider when evaluating how fast a token is going up? It could have been increasing for days or just a few minutes. We call this parameter memory length.
When you want to act, how big should that action be? Go all in or slightly increase weight of that token. We call this parameter: aggressiveness (signified by the letter k)
The below shows performance vs HODL over a historic market cycle. This example is for a BTC/ETH/ADA/DAI pool. Each pixel being a separate choice of parameters. You can see that the strategy performed well over a large set of parameter choices.
Snippet and Monte Carlo Training
Choosing the best parameters that
Monte Carlo price performance
This is a Paragraph. Click on "Edit Text" or double click on the text box to edit the content and make sure to add any relevant information that you want to share with your visitors.
MEV protection modelling
It is very important to model what guard rails need to be put in place for a strategy to run with a certain level of multi-block MEV. These representations while easy to digest have to fix the other variables involved and only vary two.
In reality you want search all guard rail settings to find the level of protection that suites the deployed chain (faster block times require more block protections) and your comfort level. You can do this in the Robodex tooling.
Efficiency compared to CEX
A good question is, why run momentum as a AMM pool. Can you rebalance on a CEX in a more efficient manner?
The above shows historic modelling of the momentum strategy efficiency vs running the same strategy on a CEX.
This modelling includes complex fee and no-arbitrage region modelling. See the whitepaper for full details.