(This is a four-minute read)
Last week, we weighed the merits of simple dollar-cost averaging (a consistent, equally-sized investment) against throwing all your money into crypto at the same time and holding-on-for-dear-life. (You can read it here)
But if you take a mathematical, risk-adjusted look at things, neither dollar-cost averaging nor long-term holding is the best approach for investing. Today, I want to present a better framework for investing into cryptocurrencies.
Dynamic Dollar-Cost Averaging
Dynamic dollar-cost averaging using risk-adjusted levels is often brushed over in favor of more simple models. Most who’ve been in the crypto space for a while will recognize its simplistic beauty: the framework reduces risks while simultaneously boosting returns.
While there are plenty of variations on the concept, we’ll take a look at the most common one today: the rainbow chart.
The Rainbow Chart
The famous rainbow chart, created in 2014, attempts to predict the fair valuation of cryptocurrency over time using a logarithmic regression matched to existing price data.
The logarithmic regression shows us two things:
- Periods of overvaluation, where price is above the best-fit regression line, which represents an opportunity to take profits
- Periods of undervaluation, where price is below the best-fit regression line, and downside risk is low
Here’s the original chart:
Note the scaling axes, illustrating Bitcoin’s exponential growth and diminishing returns.
Later, the namesake rainbow bands were added, representing different levels of risk for buying and selling.
A chart also exists for Ethereum:
While the rainbow chart has been off by thousands of dollars at times, the original regression does have some pretty remarkable predictive value: it predicted bitcoin to cross the $100,000 mark on July 16, 2021, all while seven years ago while bitcoin was trading at about 1/100th of its current price
Given an all-time high this year of $64,000, if you have an older, more accurate model, please let me know.
No, the rainbow chart hasn’t been perfect (and one shouldn’t expect it to be), but it’s been pretty good at predicting high-overvaluation events and pushing users to capitalize on phenomenal buying opportunities.
The model doesn’t have to be perfect, it just has to identify periods where risk is higher or lower than normal.
The Way it Works
If we dive into the rainbow chart itself, the blues and greens here represent a relative undervaluation, and thus a buying opportunity. Price tends to go up after spending time at these risk levels, and we should try to maximize the amount of capital we invest at those levels.
Reds and oranges are overvaluations and represent high risk and a potential opportunity for profit-taking. If we sell at these levels, we can take profits to reinvest at the bottom of the cycle.
The middle yellow band is theoretically a ‘fair’ valuation, centered around the original regression line. It can be used as a sell signal for conservative investors, a hold signal for average investors, or a buy signal for aggressive investors.
Using the Rainbow Chart as an Investing Model
But at the end of the day, we don’t care too much about the math or logarithms or best fit lines. We care about how to make money in cryptocurrency. Let’s take a look at a real-world execution of how you an ordinary investor might take advantage of the framework.
As always, we’ve got to compare our investment to some type of benchmark. Last week, we compared an $1800 lump sum investment to a simple dollar-cost-averaging investment, deploying $100 over 18 two-week periods. Today we will add this model to the mix and see how it performs relative to these other two strategies, using the last six months as our point of reference.
To make a fair comparison to other models, we’ll try to invest as close to $1800 as possible over the course of this experiment. Here’s the strategy:
- At bottom (cool-colored) risk levels we’ll invest, adding $500 every two weeks at the lowest risk level and scaling down at higher risk levels ($400 for the second-lowest, $300 for the third-lowest, etc)
- At the fair valuation, we’ll invest $100 every two weeks
- Above the fair valuation, we’ll take profit off the table bit by bit, maximizing at the top red band, where we take out $500 per two weeks
Since the ethereum and Bitcoin rainbow charts are both widely available for free, we’ll take a look at those two investments.
The Results:
Price Action:
Below is the price action of each strategy for the two cryptos we’ve analyzed. Neither of the strategies has ever spent over $1800, note the last row where you can see the cash withdrawn from each position.
For the real data nerds, I’ll include the tables, comparing holding, simple DCAing, and this dynamic cost averaging. The last column is the input or output when using dynamic DCA. Note that both models end up with you holding some amount of cash.
The Revelations
First, a disclaimer: the rainbow chart has no predictive value in the face of a black-swan event (a global loss of internet).
And the model could completely fall apart given any of the several possible bear cases for the bitcoin (heavy government regulation, an innovation resulting in the asset becoming obsolete, another cryptocurrency massively surpassing bitcoin in terms of utility, or some security flaw in the underlying technology). But it has been useful up until now.
These ideas hardly ever get talked about and are often dismissed as FUD (fear, uncertainty, doubt) but could completely disrupt the upward growth of cryptocurrency and really mess with the investment models that work today.
Why I Like Dynamic Dollar-Cost Averaging
There are a couple of things to note with this risk-adjusted investment strategy:
- Dynamic DCA produces higher returns than simple DCA
- Dynamic DCA builds in a profit-taking strategy
- Dynamic DCA isolates risk
- Dynamic DCA produces lower returns than just buying outright at a low risk level
- Dynamic DCA minimizes the pain of dips
Why is DCAing better than just holding, then?
I think this is a fair question to ask, and an easy criticism of any DCA strategy. If you buy at a low enough price, you’ll beat any DCA strategy over time.
The difference comes when we look at volatility. Here’s a table of the biggest dips over our observed time window for each strategy:
So if we factor volatility into the picture (biggest percentage drop) two things become clear:
- Simple DCAing produces higher volatility
- Simple DCAing produces worse returns
So it makes a lot of sense to elect dynamic DCAing over simple repeated purchases. And if we weigh dynamic DCAing against buying and holding with a lump sum of capital:
- Holding produces higher volatility
- Holding produces mixed returns, depending on your luck
Are there time periods where a simple DCA might have outperformed a dynamic DCA strategy? Yes, but only at high-risk levels, where you would have continued to shovel money into your investment even though it’s close to a peak. Dynamic DCA has you pulling out profits at that point in a market cycle.
And then, inevitably it drops to lower risk levels, and you continue to throw in the same amount of capital without taking profits, while a dynamic investor raises their investment. Given an overall increase in price over time, an investor prepared to adjust based on risk outperforms less sophisticated strategies every time.
Where We Stand
Today, the Bitcoin valuation sits comfortably in the “hold” band, and I think that serves as a benchmark for overall market sentiment: investors know that Bitcoin isn’t as undervalued as it was two months ago, but many are waiting on a move up to a higher price level.
So if a dynamic DCA strategy is appealing to you, consider periodically checking those rainbow bands. For now, it’s a good time to just hold on, but equipped with this knowledge, any move to the upside or downside could be a good signal to take action.
Remember: a prepared investor is a successful investor.
Thanks for reading! This is my newsletter archive, which is mostly updated, but you can read fresh content just like this every day on my daily newsletter. Subscribe for free here and follow me on Twitter here.