- Speaker #0
Hello and welcome back to Papers with Backtest podcast. Today we're diving into another algo trading research paper.
- Speaker #1
Oh yeah.
- Speaker #0
Buckle up because we're about to explore some fascinating Bitcoin trading strategies based on a paper. All right. It's called Seasonality, Trend Following, and Mean Reversion in Bitcoin.
- Speaker #1
This paper is intriguing because it doesn't just, you know, theorize about these strategies. It actually backtests them using historical Bitcoin data. So we get to see Not just like what the rules are, but how they would have actually performed if you'd, you know, traded them.
- Speaker #0
Nice. Okay, let's unpack this a little bit. The paper explores two main types of trading strategies, trend following and mean reversion. Trend following is essentially trying to ride the wave of an upward trend. You buy when the price is going up and aim to sell when it climbs even higher.
- Speaker #1
Yeah, you could think of it like surfing, you know. Okay. Catch a rising wave. And write it as far as it'll take you. A classic example of this in the stock market is the 12-month momentum effect. Oh, interesting. So stocks that have performed well over the past year often continue to outperform in the following months.
- Speaker #0
Interesting. So how did the researchers apply this concept to Bitcoin?
- Speaker #1
They tested a very specific trend-following rule by Bitcoin when it hits its maximum price over a set period of time. They looked at periods of 10, 20, 30, 40, and 50 days.
- Speaker #0
hold it for one day then sell so essentially buy high and sell higher it sounds a bit counter intuitive right did it actually work well that's where it gets interesting this strategy actually performed pretty well particularly the
- Speaker #1
10-day uh hamac strategy okay it generated an annualized return of around 41 percent yeah and a maximum drawdown of about 42 percent okay that's significantly better than simply holding bitcoin passively over the same period okay which yielded uh about 29% return, but with a much larger 83% drawdown.
- Speaker #0
Wow. So chasing those short-term highs actually paid off in this case.
- Speaker #1
It did. But remember, past performance is not indicative of future results. Of course. And it's important to consider the risks involved.
- Speaker #0
Absolutely. Now let's move on to mean reversion. This is basically the opposite of trend following,
- Speaker #1
right? Exactly. Mean reversion is based on the idea that an asset's price tends to revert to its average over time. Okay. So you buy low when the price dips below that average and sell high when it bounces back. Makes sense. A classic example in the stock market is the short-term reversal effect. Okay. Where stocks that have dropped sharply in price often rebound quickly.
- Speaker #0
Got it. So how did they test mean reversion with Bitcoin?
- Speaker #1
They applied a similar approach. Buy Bitcoin when it hits its minimum price. Over a set look back period. Again, they tested 10, 20, 30, 40, and 50 day periods. Hold it for one day, then you sell.
- Speaker #0
So buy at the bottom and sell when it bounces back. Seems like a safer bet than trend following.
- Speaker #1
Really? I'd think so, wouldn't you? Yeah. But here's the twist. This strategy turned out to be riskier than expected. Really? Some of the look back periods, particularly the longer ones, led to drawdowns of, get this, over 80 per go.
- Speaker #0
Whoa, that's a serious drop. Why would buying at the bottom, at least in theory, be so risky?
- Speaker #1
That's a great question. And it... you know, it speaks to the unique volatility of Bitcoin. It could be because historically Bitcoin has always bounced back from large drops, but that pattern might not hold true in the future.
- Speaker #0
Interesting. So even though the mean reversion strategy was potentially more profitable, at least in the back test, it also came with a lot more risk. That's a key takeaway for our listeners.
- Speaker #1
Absolutely. And it highlights the importance of. Not just focusing on potential returns, but also considering the potential risks of any creating strategy. Right. Particularly when dealing with an asset as volatile as Bitcoin.
- Speaker #0
I think that's a great point to pause for a moment. Before we move on to the next part of the research, what stood out to you the most about these initial findings?
- Speaker #1
What I find fascinating is that both trend following and mean reversion strategies, despite being based on like opposite principles, showed some promise in the Bitcoin market. that Bitcoin's price movements might be influenced by both momentum and a tendency to revert to a certain level.
- Speaker #0
That's a really insightful observation. It seems like Bitcoin's price behavior is more nuanced than just constantly going up or crashing. Right. Which makes me even more curious to hear what comes next.
- Speaker #1
Well, you're in for a treat because in the next part of our deep dive, we'll explore how the researchers combine these two strategies and we'll also uncover some surprising results about the box time to trade bitcoin oh exciting stay tuned welcome back It's fascinating to see how the researchers took those two like seemingly contradictory strategies, trend following and mean reversion, and they combined them.
- Speaker #0
I know, right? We left off with both strategies showing some potential individually, but with our own sets of risks. Yeah. So I'm really curious to hear what happens when you kind of merge the two. Did they just like buy whenever the price moved significantly in either direction?
- Speaker #1
That's pretty much it. They focus specifically on those 10-day look back periods. where both strategies showed the most promise. Remember the 10-day max for trend following and the 10-day MIN for mean reversion.
- Speaker #0
Right, right. So they basically combined those two rules into one.
- Speaker #1
Exactly. They created a new rule. It triggers a buy when Bitcoin hits either its 10-day high or its 10-day low. Pretty simple, really.
- Speaker #0
Okay, but did it work? I mean, you know, like those opposite signals, do they cancel each other out?
- Speaker #1
That's a good point. That's something I was wondering myself. but the back-tested results were pretty impressive. Oh, really? This combined strategy achieved an annualized return of almost 99%, to be exact, 98.43%. Wow. The volatility was around 48%, and the maximum drawdown was about 38%.
- Speaker #0
Those numbers definitely sound promising. But how do they compare to simply holding Bitcoin over the same period? I mean, I always like to have that benchmark in mind.
- Speaker #1
It's a great point of comparison, and the difference is significant. The combined strategies risk return profile is much more attractive. You're getting a substantially higher return with a lower drawdown compared to just holding Bitcoin. Right. Which remember had that like 83% drawdown.
- Speaker #0
Wow. So you're essentially getting the best of both worlds, capturing both the upward trends and those mean reverting dips.
- Speaker #1
Exactly. It seems like a powerful combination, at least, you know, based on this historical data.
- Speaker #0
All right. So combining trend following and mean reversion. appears to be like a winning formula for Bitcoin. Yeah. At least historically. Is that the end of the story?
- Speaker #1
Not quite. The researchers, they didn't stop there. They went on to explore another fascinating aspect of Bitcoin trading. Seasonality.
- Speaker #0
Seasonality. I always thought that was more applicable to like traditional markets like stocks. Right. Where you have patterns like the January effect or sell in May and go away.
- Speaker #1
Those are great examples. Seasonality essentially refers to... Any recurring patterns in asset prices that happen at specific times, whether it's a certain time of day, week or year. You're right that Bitcoin is traded 247. So it's a different beast.
- Speaker #0
So did they find any seasonality in Bitcoin? It seems like it would be less likely given it's always on nature.
- Speaker #1
Well, that's the surprising part. They did find something quite interesting. Okay. According to their analysis, the best time to hold Bitcoin, at least historically, was between... 21.00 and 23.00 UTC plus zero.
- Speaker #0
Wait a minute. Isn't that when most major global markets are closed?
- Speaker #1
Exactly. It does seem a bit counterintuitive, doesn't it?
- Speaker #0
For sure. So what could possibly be going on during those specific hours that might explain this pattern? Any ideas?
- Speaker #1
That's the million dollar question. And unfortunately, the paper doesn't offer a definitive answer. It's definitely a point that warrants further research. You know, it really makes you wonder. What hidden dynamics might be at play in the Bitcoin market when, you know, the traditional forces are taking a break?
- Speaker #0
Maybe that's when like the big institutional players who are less bound by traditional market hours step in. Or perhaps it's driven by like retail investors in different time zones where those hours are actually prime time for trading.
- Speaker #1
Those are some very plausible hypotheses. Another possibility is that it could be related to like specific events or news releases that tend to occur during those off-peak hours. It's really, it's a mystery that begs for further investigation.
- Speaker #0
Yeah. You know me. I love a good mystery. Especially when it comes to the world of trading. It's like we've stumbled upon a hidden clue that might unlock a deeper understanding of Bitcoin's price behavior.
- Speaker #1
And that's what makes this research so compelling. It opens up more questions than it answers, which is a sign of truly thought-provoking work.
- Speaker #0
Yeah. Before we get too lost in the maze of possibilities, let's recap the key takeaways from this research paper on Bitcoin trading strategies.
- Speaker #1
Sounds good. It's always helpful to summarize what we've learned, especially with a paper as dense as this one.
- Speaker #0
OK, so let's take a step back and kind of like recap what we've learned from this deep dive into Bitcoin trading strategies. It's been quite a journey.
- Speaker #1
It has.
- Speaker #0
We started by looking at trend following. and mean reversion strategies.
- Speaker #1
Yeah.
- Speaker #0
Both of which showed some potential for profit when applied to Bitcoin. And what was interesting is that the trend following strategy, the one that involved buying at recent highs, actually performed better than we kind of expected. While buying at recent lows, which you'd think would be safer, turned out to be quite risky.
- Speaker #1
Exactly. That was a very surprising finding and it really highlights the unique nature of Bitcoin and how its price movements can kind of defy traditional market logic.
- Speaker #0
Then we saw that combining these two seemingly opposite strategies, buying at both highs and lows, led to even better results. The backtested returns were impressive. And the risk return profile was significantly better than simply holding Bitcoin.
- Speaker #1
That's right. The combined strategies seem to effectively capture both the upward trends and those mean reverting dips. It's almost as if it was like designed for Bitcoin's volatile nature.
- Speaker #0
And then just when we thought we'd seen it, all the researchers threw in that curveball about seasonality. Right. Who would guess that the best time to hold Bitcoin might be when most traditional markets are closed?
- Speaker #1
I know, right? It's a real head scratcher. And the paper doesn't unfortunately provide a clear answer as to why this might be the case.
- Speaker #0
So what does this all mean for our listeners? Yeah. They're probably like eager to apply this knowledge to their own trading.
- Speaker #1
Well, first and foremost, it's important to remember that, you know, this is just one research paper. Of course. And past performance is never a guarantee of future results. But it does provide some valuable insights into the potential of algo trading strategies for Bitcoin and how these strategies can be tested and evaluated.
- Speaker #0
Absolutely. It highlights the importance of looking beyond simple buy and hold strategies and exploring different approaches that might be better suited to Bitcoin's unique characteristics. Right. And it also underscores the need to carefully consider both the potential rewards and risks of any trading strategy before implementing it.
- Speaker #1
And perhaps most importantly, it encourages us to keep asking questions and to never stop learning. There's clearly a lot we still don't know about the Bitcoin market and how it behaves. For sure. What drives those late night price surges? Yes. Why do certain strategies work better than others? These are all questions, you know, worth exploring.
- Speaker #0
Exactly. And that's what we love about this show. Right. It's not just about presenting information. Yeah. It's about sparking curiosity. Right. And encouraging our listeners to kind of dive deeper into the world of algo trading and to think critically about the strategies and techniques they're using.
- Speaker #1
So we'll leave you with this final thought. If the best time to trade Bitcoin is when the rest of the financial world is asleep, what does that tell us about the nature of this emerging asset class? Is it truly a 247 market driven by forces beyond traditional finance? Something to ponder.
- Speaker #0
Thank you for tuning in to Papers with Backtest podcast. We hope today's episode gave you useful insights. Join us next time as we break down more research. And for more papers and backtests, find us at https.paperswithbacktest.com. Happy trading.