# 10000 Waves Strategy

**Summary:** 10000 Waves is purely mechanical trend following strategy, trading ETH **and** BTC vs USDC. The main design goals are:
* making no assumption whether BTC or ETH will perform better in the future
* drawdown prevention while maintaing high profitability
* overfitting mitigation.

**How does it work:** 10000 Waves Strategy works in two steps. In step one it checks whether ETH is going up or down versus BTC. If ETH/BTC ratio is rising, it "allocates" more to ETH, if not, it "allocates" more to BTC. I used parenthesis, because at this step no actual trading takes place. Step one creates artificial "fund" which is sometimes 100% in ETH, sometimes 100% in BTC, but most of the time somewhere in between.

Then, in step two, it compares this "fund" performance to USDC. If "fund" dollar value goes up, it allocates more to the fund (= ETH, BTC or mix of those two). When crypto markets are bearish, it stays mostly or even solely in USDC.

Step two ensures that strategy allocates to cryptos only in bullish market conditions, while step one gives additional performance boost, ensuring that it buys better performing coin.

As opposed to many other strategies, 10000 Waves isn't switching between holding 0% to holding 100% of some asset overnight. It rather buys/sells cryptos gradually. The advantage of such approach is that it tends to trade lesser amounts so it's less affected by slippage and other trading costs.

**Drawdown prevention:** When comparing strategies, especially those dealing with cryptocurriences, it's not enough to look at profits. Simple Buy and Hold strategy ("Hodling") can deliver lots of profit, given that many cryptos experienced wild rides upwards. Drawdown prevention is much more important, and even many holders will admit that they would prefer selling ETH in early 2018 rather than seeing it losing it more than 93% of value. 10000 Waves Strategy is designed to prevent this painful situation and it goes to holding USDC when market conditions are rough. It may mean that it may miss some move up, but in the end backtests show that it manages to slash drawdown by half.

**Overfitting:** I've seen (and even developed myself) many strategies which gave phenomenal results in backtests. Most of them were overfitted: parameters were selected in such way, that strategy made some good trades and wound up very profitable. Modern software (even Excel) make it pretty easy to optimize strategy by trying all combinations of parameters and then choosing parameters that give the best results. The problem is that markets do not work that way. There's no magic indicator with some "proven" parameter that guarantees working in the future.

That's why I tried at least mitigate overfitting by **dividing strategy into 100 substrategies**, each with different parameters. This pertains to step one and step two, hence the strategy name: 100 * 100 = 10000 Waves. In both steps, strategy add up results of 100 substrategies.

I could easily pick the best substrategy for step one and then find out which substrategy for step two yields the best results. Probably I could slash drawdown even more and I'm very confident I could show some extraordinary profits. I decided not to do this, not only because it's dishonest but also because I prefer smaller profits that are real than some big amount of money that will never materialize.

**How the strategy was developed and backtested**
The strategy is an improvement over simpler strategy trading ETH vs USDT on centralized exchange. I'm developing strategies using Amibroker (dedicated software for strategy development), but in this case, since it's "strategy of strategies" I've also build my own implementation of backtester from scratch and automated it using Node.js.

Trading costs (fees and slippage), which are very high on crypto markets, can easily eat your profits, especially if the strategy trades frequently. That's why in all tests I've included 0.3% of trading costs. This is quite high and pessimistic, but

**Performance, risks and your expectations:** No strategy is 100% free from overfitting and no strategy is 100% free from assumptions. Substrategies at each of two steps operate within some range of parameters that may not be optimal in the future. There's also a risk that crypto markets changes in some deep, radical way and render this (and probably also other) strategy useless.

So it's quite important to understand assumptions on which 10000 Waves is built in. There's one fundamental assumption: that both ETH/BTC and "crypto"/USD markets are prone to trend for quite long time (if you don't believe this will be the case, perhaps you should look into other strategies). The longer, smoother the trends – the better.

For bullish markets, 10000 Waves will try to jump onto "faster horse" (ETH or BTC). It may not catch the whole upwards movement (on the plus side it may sometimes make additional profit during longer pushbacks and save you from sleepless nights), so you can't expect it beating "Hodling".

Bearish markets are where this strategy really shines. It's not addicted to hopium and it's not trying to be "tough", holding cryptos or trying to "catch the falling knife". We still live in fiat world, and USD is safe haven short and medium term.

Sideways periods are mixed bags and it depends a lot how some particular maker looks like (sometimes it's flat and sometimes very bumpy). 10000 Waves tries to squeeze some profit from medium term bull runs, and in 2019 it was quit successful at it.

Last but not least: this is long term investment strategy (if you're after profits) or at least medium term (if you're after risk mitigation). No magic will happen in two or three or even six months and underperformance is quite possible. Two years or more is optimal holding period.

**Why choose 10000 Waves**
I'm well aware choosing a strategy is hard. What's even harder is chosing a strategy which has "worse" performance than some competing strategies. Why one could even do this? Because what's matters is future, not past profits. I'm not selling you a lottery numbers that once won billions.