Everyone hates uncertainty and losing money. Backtesting is a process that removes both factors from the equation while simultaneously improving your trading skills. But although I have explained backtesting in my previous article, you haven’t explored all the nooks and crannies.
There are a lot of tidbits that make backtesting efficient, but there are equally numerous mistakes to avoid. In this article, I’ll make your life easier by giving you 5 pro tips for backtesting strategies in crypto. Who knows, one tip could be enough to change your entire backtesting game.
Backtesting is a process in which traders compare a strategy against historical market data to determine its efficacy. You select a timeframe and apply the strategy to a crypto trading pair. In return, you receive an overview of how your strategy would have fared in the past. If successful, a strategy that performed well in the past might again do so in the future.
When trading, you might have more than a few strategies in mind. Since your goal is to maximize capital protection and increase your risk/reward ratio, it’s important to pick the strategy with the best performance. Failing to do so may lead to unnecessary losses.
There are two ways to backtest crypto trading strategies. You compare a strategy against market data either manually or automatically. Both options have their benefits, but only one fits your goals and abilities. So how does each backtesting method work in practice?
Manual backtesting works by paper trading an asset on a selected timeframe. Paper trading means writing down buy and sell orders and checking the results of your imaginary trading activities. Once you collect data for a certain time period, you’ll see the strategy’s performance.
Automated backtesting works the same. You still select a timeframe, trading pair, and the strategy. The difference is that a tool automatically collects data.You don’t have to write down every order and calculate losses and profits. A trading software does this for you, saving precious time.
Which backtesting method do I recommend? It depends on your time and experience. If you don’t have enough time, manual backtesting is a big no no. Creating a spreadsheet, inputting and creating data takes too long. You should use an automated backtesting software instead. However, you’ll need the experience to use it. You don’t want to get false results from an automated backtest.
Do you like to cook? If so, you might have noticed that your dish rarely turns the way your favorite three-star Michelin chef makes it. You follow each step of the recipe but the texture, flavor, or appearance isn’t quite like the original. In both cooking and trading, you’ll find that the most important parts aren’t always learned from a recipe or article – they’re learned from experience.
To save you precious time and even more precious money, I’m giving you 5 tips for backtesting crypto strategies.
One mistake I’ve noticed in my own backtesting experiments is that I’m very selective with market data. I’ll scour through BTC/USDT or ETH/USDT and select a timeframe that I believe would work well with my strategy. I turn out to be right and the strategy performs exceedingly well. But once I trade with real money, I discover that it’s not as efficient as I thought.
Maybe my luck is bad, but I then decide to backtest the strategy again. And guess what? I discover that the strategy is useless in almost all other trading environments and timeframes. This mistake is too costly, so instead, I try to backtest random market data.
When you select a timeframe for your backtest, make sure it’s a random one. You don’t want your bias to influence results because in real trading, you won’t see the market structure clearly until it already forms weeks later. This is another excuse to use even more data in your backtest, so that your results stay objective and consistent across all timeframes.
When buying a tech product I tend to look at as many reviews as I can. I confirm if the product is good by checking multiple sources and ratings. I’ll look at reviews from different stores, watch videos, and even read someone’s 10-minute rant on a niche forum.
What happens many times is that by analyzing enough data, I discover hidden flaws. The average user won’t notice, but a few experts will find the small yet off-putting feature – usually annoying enough to trigger your hidden OCD.
I happen to utilize the same strategy when backtesting. I’ll backtest a strategy as much as I can with as much data as I can find. And more often than not, I’ll find the weakness in my strategy. But just because I find a flaw doesn’t mean that the strategy is useless. The newfound knowledge helps me correct my strategy and make it efficient.
If you have the time, you don’t have to worry too much about backtesting. You should feel like you haven’t tested the strategy enough. Only after getting sick of analyzing data will you feel confident in your choice.
We all create a strategy that sounds good on paper and instantly think we’re a genius with undiscovered talent for sitting at home and making money. But it’s just one strategy out of many. You shouldn’t marry it in the heat of the moment. There are too many good strategies in the wilderness to stick to one.
When you backtest, remember to experiment with multiple strategies. If you discover a strategy that performs well, there might be another strategy that performs better. And considering that you should minimize risk and maximize profits, even a 3% difference in performance is enormous.
What I like doing is thinking of at least five strategies and backtesting each one. Once I receive my results, I compare the strategies and select the best one. By doing so, I ensure that I stay at the top of my game and avoid unnecessary losses.
Which backtesting metrics should you care about? Is profiting more important to you than avoiding losses? If so, you should target strategies with higher-than-average returns. But if you favor a different metric, then the other data-type should attract most of your attention.
When backtesting, the following types of data come up:
If you’re trading long-term, a metric such as annual profits is more important to you than daily profits. Or if you’re interested in profiting consistently, you might not care that much about volatility. Instead, average profits might be more intriguing. Determine important backtest metrics and you’ll understand the specific usefulness of your strategy.
I might have given the impression that you should chase perfection. But you shouldn’t aim for perfection when backtesting. What backtesting results provide you with is an average overview of how a strategy would have fared at any given moment. But is a consistently good strategy really what you need?
Most traders make their money in unusual trading environments. No one profits in a boring time period where you can anticipate everything. Traders have made a lot of money during the March crash, during regulatory announcements, during China’s crypto bans, and during Ethereum’s numerous network delays. A backtest won’t tell you if a strategy will perform during these unexpected moments because it’s not designed to predict them. It’s designed to analyze the average and common market environments.
You shouldn’t be too extreme with your requirements. If a strategy is great, it means that it’s great across the average duration of the timeframe it was tested in. A strategy with a lower score might be worse on average, but it might perform better during unexpected moments when you can make the most money.
Backtesting is the process of analyzing a strategy’s performance. The most important thing about backtesting is collecting huge amounts of data and putting it to good use. However, you shouldn’t be too critical about backtesting results. Just because a strategy has a positive or negative result doesn’t mean that it will be bad in the future. You’ll only know the strategy is successful after trading crypto in real-time.
The bottom line is that backtesting can have a huge impact on your trading success. When used properly, it can help you find hidden flaws in a strategy. You can use this insight to drop the strategy or to correct it until it’s profitable. But masterful backtesting requires experience which provides you with the wisdom needed to know what data to look for and what data to avoid.
If you want to learn more about backtesting, I recommend reading the following articles:
Each day Shrimpy executes over 200,000 automated trades on behalf of our investor community. And joining them is easy.
After you sign up and connect your first exchange account, you’ll deploy an investment-maximizing strategy in as few as 5-minutes.
Whether you create your own rebalancing strategy or completely custom automation, the ability to walk your own path belongs in the hands of every crypto investor.
Discover 5 popular alternatives to Robinhood and decide which app you should start investing with in 2023.
Dive into automated crypto investing. Learn how it works, its advantages, potential risks, and top platforms for your investment journey.
A crypto robo-advisor is a platform that manages your cryptocurrency portfolio automatically. Here are 5 robo-advisors to use in 2023.