Our last case study showed that it pays to rebalance a diversified portfolio in a bull market.
All periodic rebalancing frequencies outperformed HODL in our last study.
Even our least efficient strategy beat HODL by 5.44%
But many of you know that periodic rebalancing isn’t the only rebalancing strategy out there.
The second most popular rebalancing strategy is threshold rebalancing.
To show you the difference in performance between these two rebalancing strategies, we have decided to analyze how different our last case study would have been had we applied threshold rebalancing to our backtests.
Not sure how threshold rebalancing works? Let’s have a quick reminder first.
This case study was made with the help of Shrimpy's native backtesting solution. You can use Shrimpy to analyze thousands of cryptocurrencies and compare their performance against different strategies. Backtesting is the easiest way to analyze the potential performance of your portfolio without having to deploy real capital.
Threshold rebalancing and periodic rebalancing are the two most popular rebalancing strategies. Both are fairly easy to understand. While one rebalances a portfolio on a time-basis (every hour, day, week, or month), the other uses tolerance bands to keep target asset allocations in check.
The following two sections provide you with an in-depth overview of how each strategy works.
Simple illustration depicting how periodic rebalancing takes place at specific times. After 24 hours, the allocations are not equal, so a rebalance will make them equal once again.
The simplest rebalancing strategy is periodic rebalancing, which uses a fixed amount of time between each rebalance. This amount of time is usually shorter for cryptocurrencies than for other asset classes due to rapid price fluctuations. For example, it would be reasonable to select a portfolio rebalancing frequency of 1 day. This would mean that at the same time every day, your portfolio would be rebalanced.
The most used rebalancing frequencies include:
A demonstration of threshold rebalancing when a portfolio reaches a 20% deviation from its target allocations. Notice the green and the blue assets are 20% away from their target allocation of 25%. This difference between the target and current allocations triggers a rebalance.
Rebalancing based on allocation tolerance bands examines the drift of the allocations relative to their target or desired allocations. By evaluating this difference, the drift of each asset is tracked over time. So as the percent allocation of individual coins drifts further from the desired percentages, a rebalance takes place when the difference between the current and target allocations crosses a threshold.
For example, with a threshold that is ±20% on an asset with a 25% target allocation, if the asset represents 5% more or less of the entire portfolio than this target, the portfolio is rebalanced.
Imagine the situation previously discussed where we had 4 different coins that each held 25% of the portfolio value. In this method, a rebalance would happen as soon as one of those assets consumes less than 20% or more than 30% of the portfolio value. However, this also means that if all of the coins in the portfolio are increasing or decreasing in value together – without changing their percent representation in the total portfolio – then no rebalance takes place.
This case study focuses on the impact of threshold rebalancing on cryptocurrency portfolio performance during a 3-year time period. The time period includes the 2020-2021 bull market, as well as the 2022 bear market.
We've analyzed 7 settings based on threshold rebalancing tolerance bands. We've run 7,000 backtests in total, amounting to 1,000 backtests per setting.
This time we’ve only analyzed a 10-asset portfolio.
Our threshold rebalancing includes the following tolerance bands:
This case study compares the performance of each rebalancing strategy against HODL (buy-and-hold). We've also compared the performance of each strategy against the initial portfolio value.
Note that the asset selection process during our backtests was completely randomized.
The case study includes one portfolio group in terms of portfolio size. The portfolio has 10 assets in total.
The assets in the portfolio group are evenly distributed. In this case, the portfolio has 10 assets with a 10% allocation per asset.
The rebalancing process simulated in the backtest executes trades to keep these initial allocation targets in check. The point of rebalancing is to maintain targeted allocations.
Note that each executed trade includes a 0.1% trading fee.
Each portfolio starts with an initial balance of $5,000.
The results of this case study show the value of each portfolio by the end of the backtest period (January 1st, 2020 - December 31st, 2022.)
The asset selection process in this case study is completely randomized for the purposes of objectivity and accuracy.
Our list of assets includes all cryptocurrencies that were available on the top 10 crypto exchanges during the backtest period.
Each backtest picked random assets for each portfolio group based on the list of assets available at that time. This is done so that the focus of our analysis is placed not on the assets themselves but on the strategy.
Each backtest outputs two results. One result shows a portfolio's final value if it had been rebalanced, while the other shows a portfolio's final value if it had used the HODL strategy.
To determine how these strategies compare, we calculate the performance of the rebalancing strategy against the HODL strategy by using the following formula:
Performance = ((R - H) / H) x 100
You can read the formula in the following way:
Our results contain graphs that show the relationship between the number of assets held in a portfolio with the median value of the portfolio by the end of the backtesting range. We have results for HODL, monthly rebalancing, weekly rebalancing, daily rebalancing, and hourly rebalancing.
Our final results (Portfolio Rebalancing vs. HODL) compare portfolio rebalancing against the HODL strategy. This section focuses not on the median value of each portfolio by the end of the backtest, but on the performance of a rebalanced portfolio compared to the same portfolio utilizing a HODL strategy.
Please note that any values displayed in this section are not relative to the starting value of a portfolio but to the same portfolio had it been HODL’ed. If a value of 5% is displayed, that means the final result for the rebalanced portfolio is 5% higher than the HODLed portfolio, and not 5% higher than the initial portfolio fund.
Our case study focuses on the following market period: Jan 1, 2020 - Dec 31st, 2022.
Given the nature of the market during that time period, our case study predominantly analyzes the price data of a bull market.
.png)
The market has seen extremely volatile price action during this backtest period. Bitcoin has seen a price increase of 851% from Jan 1st, 2020, to the November ATH of 2021. The market has also seen a 76% price decrease from the ATH to Dec 31st, 2022.
Considering that our case study involves mostly altcoins (due to the nature of diversification), the results are far more volatile.
Our results start by individually analyzing the effects of each threshold rebalancing tolerance band performance versus a buy-and-hold (HODL) strategy.
The list of tolerance bands includes:
Our combined results section compares the combined performance of all the strategies listed above. Note that the performance in the combined results section is denominated in dollars. Their performance is relative to the initial portfolio value.
Our final results compare the combined performance of each rebalancing strategy against HODL. We have also included results from the previous periodic rebalancing case study.
The values in the final results section are percentages, not dollars. The values are relative to a portfolio utilizing HODL – the performance difference between a rebalanced portfolio and a HODL’ed portfolio – and not to the initial value of a portfolio.
The above histogram compares the percent performance of a 1% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 316 (or 31.6%) backtests produced performance results that ranged between 0% and 80% for the 1% threshold rebalancing strategy.
The median performance versus HODL is: 11.49%
The above histogram compares the percent performance of a 5% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 498 (or 49.8%) backtests produced performance results that ranged between 0% and 170% for the 5% threshold rebalancing strategy.
The median performance versus HODL is: 52.2%
The above histogram compares the percent performance of a 10% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 272 (or 27.2%) backtests produced performance results that ranged between 0% and 74% for the 10% threshold rebalancing strategy.
The median performance versus HODL is: 68.6%
The above histogram compares the percent performance of a 15% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 363 (or 36.3%) backtests produced performance results that ranged between 0% and 110% for the 15% threshold rebalancing strategy.
The median performance versus HODL is: 77.1%
The above histogram compares the percent performance of a 20% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 323 (or 32.3%) backtests produced performance results that ranged between 0% and 96% for the 20% threshold rebalancing strategy.
The median performance versus HODL is: 73.1%
The above histogram compares the percent performance of a 25% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 338 (or 33.8%) backtests produced performance results that ranged between 0% and 84% for the 25% threshold rebalancing strategy.
The median performance versus HODL is: 71.5%
The above histogram compares the percent performance of a 30% threshold rebalancing to a buy-and-hold (HODL) strategy. 1,000 backtests were included in the histogram. After each backtest was run, it was placed into one of the performance buckets to create the curve you see.
The x-axis represents the performance increase over buy and hold for each backtest in percentages. The y-axis is the number of backtests that fall into each of the performance ranges on the x-axis.
The process for constructing this histogram is as follows: once a backtest is complete, the portfolio value of the rebalanced portfolio is compared to that of the HODLed portfolio using the methodology outlined in the section titled “Performance Calculation.”
From this comparison, we receive a percentage that represents how much better or worse rebalancing has performed than HODL. Based on this percent, we increment the count for the number of backtests which fell into the corresponding performance range.
About 377 (or 37.7%) backtests produced performance results that ranged between 0% and 93% for the 30% threshold rebalancing strategy.
The median performance versus HODL is: 67.4%
.png)
This graph shows the combined results of all the previous strategies for a $5,000 portfolio after three years. The X-axis shows the tolerance band for each portfolio group. The Y-axis shows the portfolio's final dollar value at the end of the backtest period.
Our combined results show that the 15% tolerance band brings the highest returns.
Higher tolerance bands have a clear positive impact on portfolio performance, up to 15% bands. Tolerance bands higher than 15% have a negative impact on performance.
The case study indicates that increasing your tolerance band from 1% to 5% would have resulted in 35.59% higher returns. Increasing the band from 1% to 15% resulted in 64.68% higher returns.
Although the jump in performance between the 1% tolerance band and the remaining ones is incredibly significant, the results show that further jumps in performance are gradual and predictable.
The image above shows the performance of each strategy against the initial portfolio value. It’s clear that threshold rebalancing completely outperforms HODL.
The image above shows the performance of each rebalancing frequency for a periodic rebalancing strategy against the initial portfolio value. The data above stems from the previous backtest study linked at the start of the case study. Both case studies use the same backtest time period and portfolio value.
The average performance of all periodic rebalancing strategies equals: 155%
The average performance of all threshold rebalancing strategies equals: 251.06%
Our results indicate that threshold rebalancing is far superior to periodic rebalancing for 10-asset portfolios.
.png)
The graph above compares how well each threshold rebalancing strategy has performed against HODL at varying tolerance bands. The X-axis shows the tolerance band utilized. The Y-axis shows the performance, defined in percentages, of a threshold rebalancing strategy compared to a HODL strategy at the end of the backtest period.
Our results show that investors receive parabolic increases in returns by increasing their tolerance bands from 1% to 15%. However, it’s noticeable that increasing the tolerance band past 15% is inefficient and leads to lower returns.
The above table compares the performance of a threshold rebalancing strategy against HODL at varying tolerance bands.
The above table compares the performance of a periodic rebalancing strategy against HODL at varying rebalancing frequencies.
The average performance of all periodic rebalancing strategies (versus HODL) equals: 20.86%
The average performance of all threshold rebalancing strategies (versus HODL) equals: 60.20%
Threshold rebalancing is three times more efficient than HODL.
The results of our latest case study shows that threshold rebalancing has, by far, the best performance against HODL.
Threshold rebalancing also brings better returns (compared to initial portfolio value) than both HODL and periodic rebalancing.
We conclude that the 15% tolerance band brings the highest returns in a threshold rebalancing strategy. Tolerance bands past 15% bring negative results.
We have also concluded that threshold rebalancing is three times more efficient than periodic rebalancing.
Keep in mind that we only considered data for 10-asset portfolios in this case study. The results between periodic and threshold rebalancing might be different when other portfolio sizes are chosen.
The backtests and rebalancing strategies were carried out using our very own portfolio rebalancing tool, Shrimpy.
Shrimpy is an automated portfolio management platform that helps you not only rebalance but also diversify your crypto portfolio.
You can connect over 25 exchange accounts and wallets with Shrimpy and start rebalancing your portfolio right away. Shrimpy is one of the easiest to use rebalancing tools in the industry.
Sign up now by clicking here.
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.
Threshold rebalancing is a popular crypto strategy that involves a high trading activity. Find out how different trading fees affect the performance of your threshold strategy.
.jpg)
Diversification and rebalancing have a positive impact on portfolio performance. Discover which portfolio sizes and rebalancing periods are the best.
Want to optimize your threshold rebalancing strategy? Find out which crypto exchanges offer the best performance based on their trading fees!