tldr: skip ahead to the conclusion at the bottom.

note: I will not be doing TA in this post. Instead, I will be trying to use data to objectively assess how performant this popular TA indicator has been across its history (not very performant).

Introduction

Since the death cross printed on the BTC chart a couple days ago, a lot of posts have been popping up around the topic, speculating on whether this dreaded indicator will actually herald in the kind of bearish downturn that it's associated with. The prevailing sentiment on this sub seems to be that crosses like the death cross and the golden cross are basically just financial astrology, but I wanted to take a look at the actual data to try to get an unbiased perspective on the validity of this indicator.

For anyone who doesn't know, a death cross is simply when the 50 day moving average falls below the 200 day moving average, and a golden cross is when the 50 day moving average rises above the 200 day moving average. They are lagging momentum indicators that have been used for a long time in the stock market. Even there, they are somewhat dubious, but we should expect them to be less accurate the more volatile a market is, since high volatility means momentum is less reliable. A death cross really just says that the past 50 days have just now become slightly worse on average than the last 200 days were, and a golden cross just says that the past 50 days have just now become slightly better on average than the past 200 days were. These are interpreted as indicators of long-term bearish periods to come (or bullish, for golden crosses), since they suggest trend reversals on a large time scale.

The Data

For each of the 8 death crosses since 2014, I have checked the price at the time of the cross, and then checked how that related to the price 3 months later, 6 months later, 1 year later, and 1.5 years later. I chose these reference points because death crosses are meant to indicate long-term price trends, but going beyond 1.5 years would start to yield less meaningful results since crypto bull markets themselves only tend to be ~1.5 years long.

I have then tagged each death cross with an overall rating for how bullish or bearish the 1.5 years following the cross were by aggregating the 3 month, 6 month, 1 year, and 1.5 year price changes (I considered the 1.5 year price change to be more significant than the other 3, since it represents lasting price change over the period).

Finally, I have used this rating to determine how accurate each of the 8 death crosses have been as indicators of long-term bearishness to come.

^\note: some of the numbers here might be inaccurate by maybe) ^{±1 week because I used a very large timescale chart for this data, but I don't think it should matter too much in the end.\})

April 2014 Death Cross

• Price: $360
• 3 month change: +71%
• 6 month change: +0%
• 1 year change: -32%
• 1.5 year change: -24%
• Overall rating: bearish
• Indicator accuracy: accurate

September 2014 Death Cross

• Price: $480
• 3 month change: -31%
• 6 month change: -42%
• 1 year change: -50%
• 1.5 year change: -11%
• Overall rating: bearish
• Indicator accuracy: accurate

September 2015 Death Cross

• Price: $235
• 3 month change: +88%
• 6 month change: +79%
• 1 year change: +162%
• 1.5 year change: +377%
• Overall rating: very bullish
• Indicator accuracy: very inaccurate

March 2018 Death Cross

• Price: $6,950
• 3 month change: -1%
• 6 month change: -4%
• 1 year change: -44%
• 1.5 year change: +44%
• Overall rating: neutral
• Indicator accuracy: somewhat inaccurate

October 2019 Death Cross

• Price: $8,950
• 3 month change: -6%
• 6 month change: -22%
• 1 year change: +29%
• 1.5 year change: +542%
• Overall rating: bullish
• Indicator accuracy: inaccurate

March 2020 Death Cross

• Price: $6,770
• 3 month change: +39%
• 6 month change: +61%
• 1 year change: +742%
• 1.5 year change: +579%
• Overall rating: very bullish
• Indicator accuracy: very inaccurate

June 2021 Death Cross

• Price: $37,000
• 3 month change: +27%
• 6 month change: +30%
• 1 year change: -41%
• 1.5 year change: -55%
• Overall rating: bearish
• Indicator accuracy: accurate

January 2022 Death Cross

• Price: $43,000
• 3 month change: -5%
• 6 month change: -50%
• 1 year change: -52%
• 1.5 year change: -29%
• Overall rating: bearish
• Indicator accuracy: accurate

Conclusion

The final tally is:

• Very Accurate: 0
• Accurate: 4
• Somewhat Accurate: 0
• -----------------------------------
• Somewhat Inaccurate: 1
• Inaccurate: 1
• Very Inaccurate: 2

According to these results, accuracy seems about 50/50, suggesting that death crosses are no better at predicting long term bearish trends than coin tosses are.

That statement is probably a little too bold to make based upon the analysis I have done; there is definitely some subjectivity in this data. Trends can happen on many different time scales; I tried to pick a reasonable set of time scales, but it's true that changing those numbers around would shake up the data somewhat. It's also hard to try to use a single word to qualify a 1.5 year period based on 4 data points. It's a subjective heuristic.

But nevertheless I think I can confidently say that Bitcoin death crosses are not a reliable predictor of long-term bearish trends. They frequently occur before both bearish and bullish trends.

Thanks for reading!