[r/cc Crosspost] A few weeks ago I downloaded all historical data from the Fear and Greed Index, and wrote a backtesting script in python. I wanted to find out how DCA performed if you increased your DCA amount the more FEARFUL the market was, and invested very little during times of greed. The post is here. (Yes we know the meticulous definition of DCA is keeping the amounts the same, but people usually vary the amounts regardless due to variable salaries and expenses).

Tons of people asked me to incorporate selling during EXTREME GREED, and also pointed out other metrics they wanted me to test. So here it is! [Results at the Bottom

Start date: March 17th 2018
End Date: September 9th 2023 (Bottom or Top is not a fair end date)

1. Benchmark DCA

Traditional “vanilla” DCA to measure our results against.

• $100 invested each 7 days

2. Fear & Greed DCA NO SELLING

Each 7 days we check the F&G index and invest:

• $150 if “Extreme Fear” $100 if “Fear” $75 if “Neutral” $50 if “Greed” $25 if “Extreme Greed”

3. Fear & Greed DCA WITH SELLING

Each 7 days we check the F&G index and:

• $150 if “Extreme Fear” $100 if “Fear” $75 if “Neutral” $50 if “Greed” Sell 5% of accumulated BTC if “Extreme Greed”

4. PI Cycle Top Indicator

More than one person asked me to test this one. The PI cycle top is the 111D and the 350D MA x2. It can be found here for example. The idea is when the 111D cross above the 350Dx2, you should run for the hills.
Each 7 days we check if the 111D is below the 350Dx2 and:

• $100 invested if below Sell everything if above

5. AΙpha-Squared Risk Metric DCA

The top comment mentioned this one. I signed up for a free account to test it so you don’t have to. I used their default “Aggressive” strategy with 0% profits reinvested.

Each 7 days we check the risk and:

• $130 if risk if 0/100 $120 if risk is 20/100 $50 if risk is 55/100….. And so on Sell 30% if risk 80/100 Sell 70% of if risk 90/100

6. Drumroll.......The results!

\ Final Value (BTC) = The value of your entire portfolio in BTC. We value like this since even for the strategies where you sell to fiat, you can immediately convert it to BTC at the final price if you like.*
\ SATS per dollar = Wealth-in-BTC / Amount-Invested*

IMPORTANT DISCLAIMERS!!

• you NEED to test this on multiple timeframes. This is NOT conclusive evidence. I'm just a simple data engineer who likes to explore stuff like this.
• F&G is NOT a super accurate metric on longer timeframes, hence why I picked weekly DCA.
• PI Cycle Top & AΙpha-Squared DOES CONTAIN some backfitted data for this period. However, MOST of the data is NOT backfitted, I made sure to check. So the results are still fairly indicative.
• DCA in GENERALLY UNDERSTOOD as buying recurrently regardless of the asset price. Everyone is well aware that buying for slightly different amounts is not technically speaking DCA, but very few people can follow this rigid definition in reality.
• Finally, don't take this too seriously :) It's the weekend and this is how I spend my free time, I just wanted to share. I'm not saying any of these strategies are better than something else. It's all for the experiment!

Regardless these are very interesting results. I hope you find it useful in the upcoming bull! I will post an update doing the same test but across multiple timeframes. I'll aim to completely exclude any backfitted data as well as try F&G on a timeframe where it performs weaker such as monthly.