Description

Are you still relying on outdated investment strategies that are likely leading you to underperformance? In this episode of "Papers With Backtest: An Algorithmic Trading Journey," the hosts dive deep into the compelling research paper titled "The Trend is Our Friend, Risk Parity, Momentum and Trend Following in Global Asset Allocation." This enlightening discussion unpacks the limitations of traditional investment approaches, such as the classic 60-40 stocks and bonds allocation, and reveals how emotional decision-making can derail even the most disciplined investors.

As the hosts explore the nuances of algorithmic trading, they emphasize the critical importance of rule-based strategies to navigate the complexities of market behavior and enhance risk-adjusted returns. By introducing innovative concepts such as risk parity—where investments are allocated based on volatility—and trend following, which involves strategically buying assets in an uptrend while selling in a downtrend, listeners will gain valuable insights into modern investment methodologies. Furthermore, the episode delves into the realm of momentum investing, where assets are ranked based on their past performance, offering a fresh perspective on how to optimize your portfolio.

Listeners will be captivated by the hosts' discussion of the paper's findings, which demonstrate that a combination of these strategies can lead to superior risk-adjusted returns compared to traditional methods. The backtest results presented in this episode reveal the effectiveness of these combined approaches across various market conditions, shedding light on why algorithmic trading is not just a trend but a necessary evolution in investment strategy.

As the episode draws to a close, the hosts summarize key takeaways that underscore the importance of strategy diversification and the accessibility of algorithmic trading. They also highlight the potential of behavioral finance in shaping trading strategies, encouraging listeners to rethink their investment paradigms. Don't miss this opportunity to elevate your understanding of algorithmic trading and discover how you can apply these insights to enhance your own trading journey. Tune in now to "Papers With Backtest" and transform your approach to investing!

Hosted by Ausha. See ausha.co/privacy-policy for more information.

Transcription

• Speaker #0

  Hello, welcome back to Papers with Backtest podcast. Today we dive into another algo trading research paper.

• Speaker #1

  The paper we're exploring is titled The Trend is Our Friend, Risk Parity, Momentum and Trend Following in Global Asset Allocation.

• Speaker #0

  So not just another basic 60-40 stocks and bonds conversation. Right. This sounds like it could get interesting for our listeners who are looking to build more sophisticated algo trading portfolios.

• Speaker #1

  Absolutely. This paper really digs into how we can potentially improve risk-adjusted returns by combining several different strategies. It's all about moving beyond that traditional approach.

• Speaker #0

  OK, I'm intrigued. But before we get into the strategies themselves, the paper starts by highlighting something a bit discouraging,

• Speaker #1

  right? It does. Even when the markets are doing well, the average investor often underperforms. And you know what? It often comes down to our own behavior.

• Speaker #0

  You're saying we're often our own worst enemies.

• Speaker #1

  Yeah.

• Speaker #0

  That's a bit unsettling. Why does that happen?

• Speaker #1

  Think about it, when markets drop, fear takes over. People panic and sell at the worst possible time. Then, on the flip side, they chase those hot stocks, buying high. It's emotional decision-making, and it rarely leads to good outcomes.

• Speaker #0

  So this paper is advocating for rule-based strategies to help us overcome these psychological pitfalls.

• Speaker #1

  Exactly. If we can take the emotion out of it and stick to a predefined set of rules, The paper argues that we can make more rational decisions and potentially improve our returns.

• Speaker #0

  OK, that makes sense. Yeah. But how do these strategies actually work in practice? Well… What are the rules we should be following?

• Speaker #1

  Let's start by addressing the limitations of that classic 60-40 portfolio. The problem is that the volatility of stocks tends to dominate, even with 40% allocated to bonds. So your portfolio is still heavily exposed to the ups and downs of the stock market.

• Speaker #0

  So even with a supposedly balanced portfolio… You're not really getting the risk reduction you might expect.

• Speaker #1

  That's where the concept of risk parity comes in. Instead of fixed percentages, risk parity allocates investments based on the inverse of their volatility. The idea is to balance risk across different asset classes.

• Speaker #0

  So if bonds are less volatile than stocks, you'd allocate a larger portion to bonds to balance things out.

• Speaker #1

  Exactly. You're essentially aiming to create a portfolio where each asset contributes equally to the overall risk. However, the paper points out that risk parity on its own doesn't take future expected returns into account.

• Speaker #0

  Interesting. So it's a good starting point for risk management, but it might not be the whole picture. What else do we need to consider?

• Speaker #1

  Well, that's where trend following comes in. This strategy is all about buying assets that are in an uptrend, meaning their price is above a certain moving average and selling or going to cash when they're in a downtrend. It's about riding the wave, but knowing when to jump off before it crashes.

• Speaker #0

  That sounds simple enough. but I imagine it's harder to execute than it sounds. We humans are naturally drawn to buying low and selling high, even though that rarely works out in practice.

• Speaker #1

  You're absolutely right. It goes against our natural instincts. And what's fascinating is that there are actual psychological factors driving these trends. For example, you have anchoring bias where people fixate on a certain price level or the disposition effect where they hold on to losers for too long and sell winners too early.

• Speaker #0

  So you're saying that all these human quirks are actually creating predictable patterns in the market?

• Speaker #1

  It certainly seems that way.

• Speaker #0

  Interesting.

• Speaker #1

  Hurting behavior also plays a role. People see others buying, they jump on the bandwagon, and that fuels the trend even further.

• Speaker #0

  And of course, there's confirmation bias, where people only see the data that supports their existing beliefs. It all adds up.

• Speaker #1

  Right. All of these biases can contribute to market trends. The key takeaway for our listeners is that But trend following requires a level of discipline that many investors simply lack. It's about sticking to the rules even when your gut is telling you to do something else.

• Speaker #0

  OK, I'm starting to see how this all fits together. So we've got risk parity to manage volatility and trend following to capture those market swings. But this paper goes even further, right? It introduces another layer, momentum.

• Speaker #1

  Yes. Momentum investing involves ranking assets based on their past performance, essentially. buying the winners and selling the losers. It's about capitalizing on the momentum that's already in play.

• Speaker #0

  But isn't that risky? I mean, those high flying stocks can come crashing down pretty quickly.

• Speaker #1

  That's true. And that's where the paper's recommendation of volatility adjusting the momentum rankings comes in. By taking volatility into account, you can potentially avoid overemphasizing assets that are just experiencing temporary price spikes.

• Speaker #0

  So it's about finding assets that have sustained upward momentum, not just A fleeting moment of glory.

• Speaker #1

  Precisely. And this is where things get really interesting. This paper's most compelling finding is that combining momentum, trend following, and volatility-adjusted weighting leads to the best risk-adjusted returns.

• Speaker #0

  So it's not about picking one strategy over another. It's about blending them together for a more robust approach.

• Speaker #1

  Exactly. Think of it this way. Momentum ensures you're in assets that are already outperforming. Trend following keeps you out of downtrends, even among those winners. And the volatility adjustment helps balance the overall portfolio.

• Speaker #0

  So you're trying to capture the upside potential while minimizing the risk. That sounds pretty powerful. But how do we know this actually works?

• Speaker #1

  Well, the paper backs up these claims with some pretty compelling backtest results. They found that trend following consistently delivered better risk-adjusted returns. compared to a simple buy and hold strategy.

• Speaker #0

  So they're not just talking theory. They have the data to support it. What kind of results are we talking about? Well,

• Speaker #1

  for example, they found that a portfolio using trend following with a 10-month signal achieved an annualized return of 8.02%, with a Sharpe ratio of 0.79. Okay. That's pretty impressive, especially when you compare it to the returns of a basic buy and hold portfolio, which often struggles to achieve consistent positive returns.

• Speaker #0

  Okay, I'm definitely seeing the potential here. What else did the back test reveal?

• Speaker #1

  They found that adding trend following to a risk parity portfolio also significantly boosted performance. And when they combined momentum, trend following, AND, volatility adjusted weighting, they saw even further improvements in risk adjusted returns.

• Speaker #0

  This has started to sound like a winning formula for algo trading. Yeah. But I'm eager to hear more about those backtest results and how these strategies perform in different market conditions.

• Speaker #1

  Absolutely. Let's dive into that next.

• Speaker #0

  So let's unpack those backtest results. Where do we start?

• Speaker #1

  One of the most interesting findings was when they applied trend following to specific asset classes. Okay. They took the MSCI World Index, which tracks developed market equities, and divided it into individual country indices. Then they applied trend following to each country separately.

• Speaker #0

  So instead of just deciding whether to be in or out of developed equities as a whole, they were making more granular decisions at the country level.

• Speaker #1

  That's right. And the results were impressive. Using a 10-month trend-following signal, They achieved an annualized return of 10.16%, with volatility of just

• Speaker #0

  9.89%. That's a sharp ratio of over one, which, as our listeners know, is considered quite good.

• Speaker #1

  Yes. It indicates strong risk-adjusted returns. Right. And the maximum drawdown for this strategy was only 16.29%, significantly lower than the drawdown for the overall MSCI World Index over the same period.

• Speaker #0

  So even though you're investing in equities, which can be quite volatile, This approach helps you manage that risk more effectively.

• Speaker #1

  Exactly. And what's interesting is that this trend-following approach wasn't limited to equities. They saw similar improvements in commodities, REITs, and even emerging markets.

• Speaker #0

  So it seems like a pretty robust strategy, not just a one-market wonder. Right. Let's dig into the combination of momentum and trend-following. How did that play out in the back tests?

• Speaker #1

  That's where it gets really interesting. Remember, momentum is about identifying assets that have been performing well recently. The challenge is that momentum can be risky, so the researchers wanted to see if adding trend following could improve things.

• Speaker #0

  It sounds like a good way to potentially ride the momentum wave, but jump off before it crashes. What did they find?

• Speaker #1

  They tested a few different approaches. One was using a trend following filter at the asset class level. Let's say you have a momentum portfolio of the top performing U.S. equity sectors. You're already picking the winners within U.S. equities.

• Speaker #0

  So you're starting with a momentum-driven selection process.

• Speaker #1

  Right. But before you actually invest in that portfolio, you'd check whether the overall U.S. equity market is in an uptrend using that 10-month moving average we talked about earlier.

• Speaker #0

  So you're essentially adding a safety net to your momentum strategy.

• Speaker #1

  Exactly. And the backtest showed that this simple addition led to a significant improvement in risk-adjusted returns. The Sharpe ratio for the momentum portfolio with the trend-following filter was 0.84 compared to just 0.001%. 0.44 for the momentum portfolio alone.

• Speaker #0

  That's almost double the risk adjusted return just by adding that extra layer of analysis.

• Speaker #1

  And the maximum drawdown was also much lower, 16.28%, compared to a whopping 56.02% for the momentum portfolio without the trend following filter.

• Speaker #0

  Those results are pretty compelling. It really highlights the importance of managing risk even when you're chasing those high momentum returns.

• Speaker #1

  Absolutely. Now, they also tested applying the trend following filter at a more granular level. Instead of checking the trend of the overall asset class, they looked at the trend of each individual asset within the momentum portfolio.

• Speaker #0

  So even if the overall U.S. equity market was trending upwards, you wouldn't buy a specific stock if it was showing a downtrend.

• Speaker #1

  Precisely. And this approach also led to improvements in risk-adjusted returns. Though the results were slightly less dramatic than using the asset class level filter.

• Speaker #0

  It sounds like there's a tradeoff between complexity and potential benefit. The more granular you get, the more analysis you need to do.

• Speaker #1

  That's a good observation. The best approach will depend on your specific goals, your resources, and your risk tolerance.

• Speaker #0

  Now, I'm curious about that flexible asset allocation strategy you mentioned earlier, the one where they didn't predetermine the asset allocation percentages.

• Speaker #1

  Ah, yes. That's a really interesting concept. They didn't decide up front how much to invest in each asset class, but let the market guide them.

• Speaker #0

  Sounds like they're putting a lot of faith in the data. Yeah. How did they actually do that?

• Speaker #1

  They ranked all 95 of the assets they were considering across all asset classes based on their volatility-adjusted momentum.

• Speaker #0

  So they were looking for the assets with the strongest recent performance. taking into account their riskiness.

• Speaker #1

  Exactly. And then they built portfolios by simply picking the top performers. Okay. They tested portfolios with different numbers of holdings, ranging from five to 50, to see how diversification played a role.

• Speaker #0

  So the momentum rankings determined both which assets to buy A&D, how much to allocate to each one.

• Speaker #1

  Right.

• Speaker #0

  Interesting. What were the results of this data-driven approach?

• Speaker #1

  The results were quite impressive. They found that portfolios with between 15 and 30 holdings consistently achieved sharp ratios above 0.8. That's better than many traditional asset allocation strategies that rely on fixed percentages.

• Speaker #0

  And what about the downside? How did those portfolios perform during market downturns?

• Speaker #1

  The maximum drawdowns for these portfolios were all below 35%, which is a reasonable level of risk given the potential for higher returns.

• Speaker #0

  That's a good point. It's all about balancing risk and reward. Now, you mentioned earlier that they added a trend following overlay to this flexible allocation strategy. That seems like it would add another layer of complexity.

• Speaker #1

  It does, but the results are really compelling. By combining the volatility adjusted momentum rankings with a 10-month trend following filter, they were able to achieve sharp ratios close to 1 with maximum drawdowns below 25%.

• Speaker #0

  Those numbers are pretty remarkable. So they're essentially letting the market pick the best assets based on recent performance. And then using trend following to manage the risk.

• Speaker #1

  That's a good way to put it. And this approach had some other interesting benefits as well. For instance, it helped to reduce the concentration risk that can arise when you only focus on momentum.

• Speaker #0

  Right. If everyone is piling into the same hot stocks, it can create a bubble that's just waiting to burst.

• Speaker #1

  Exactly. But by adding the trend following filter, they were able to diversify their holdings and avoid getting caught in those momentum traps.

• Speaker #0

  So it's like a momentum strategy with a built-in safety mechanism.

• Speaker #1

  I like that analogy. It's about being smart about how you pursue those momentum opportunities.

• Speaker #0

  Now, these backtest results are all very impressive, but I do have one question. Yeah. How did they actually measure the performance of these strategies? Did they just look at the overall returns?

• Speaker #1

  They used a variety of metrics to evaluate the performance. They did look at annualized returns, but they also calculated sharp ratios to assess the risk-adjusted returns, which is crucial for getting a more complete picture.

• Speaker #0

  Right. Because a high return doesn't mean much if you're taking on excessive risk to achieve it.

• Speaker #1

  Exactly. They also calculated maximum drawdowns to see how much the portfolios lost during those inevitable market downturns. They wanted to understand both the upside and the downside potential of each strategy.

• Speaker #0

  So they were taking a very holistic view of performance, not just focusing on the shiny returns.

• Speaker #1

  They were. They even looked at the skewness of the returns, which tells us whether the portfolios were more likely to experience big gains or big losses.

• Speaker #0

  So they were... really digging deep into the data to understand the underlying characteristics of each approach. What did they find?

• Speaker #1

  The results were consistent across all of these different metrics. The strategies that combined risk parity, momentum, and trend following consistently outperformed the more traditional approaches. And this wasn't just a one-off. The findings held up across different market conditions and time periods.

• Speaker #0

  It's always reassuring to see consistent results. It. adds to the credibility of the research. What did the researchers say about the reasons behind these outperformance? Do they have any theories?

• Speaker #1

  They did. They dedicated a whole section of the paper to analyzing the risk factors that might be driving these superior returns. They wanted to know if these strategies were truly capturing something unique or if they were just cleverly disguised exposures to known risks.

• Speaker #0

  That's a really important question. We need to understand what's driving the results if we want to apply these strategies effectively. How did they approach this analysis?

• Speaker #1

  They started by using the classic Fama-French three-factor model, which includes factors for market risk, size, and value.

• Speaker #0

  Those are considered the cornerstones of traditional asset pricing models, right?

• Speaker #1

  Right. And they added a fourth factor, the momentum factor, to capture the tendency for winners to outperform losers, which is a key element of the strategies we're discussing.

• Speaker #0

  So they were testing whether these well-established factors could explain the returns they were seeing. What did they find?

• Speaker #1

  The Fama-French factors could explain a portion of the returns, but a significant chunk remained unexplained.

• Speaker #0

  Intriguing. So, these strategies aren't simply replicating those well-known risk factors. There's something else at play.

• Speaker #1

  Exactly. They then tested a second set of risk factors, which included broader market exposures like global equities, bonds, and commodities. They also incorporated hedge fund factors to capture strategies like trend following and arbitrage.

• Speaker #0

  So they were casting a wider net, looking for any hidden exposures that might be driving the results.

• Speaker #1

  That's right. And again, they found that these additional factors could explain some of the returns, but a substantial portion remained unaccounted for.

• Speaker #0

  It seems like these strategies are tapping into something unique. something beyond those traditional risk factors.

• Speaker #1

  That's the implication. Now this is where it gets really intriguing. The paper doesn't definitively say what this something unique is, but they do offer some thought-provoking insights.

• Speaker #0

  Okay, I'm all ears. What are their theories?

• Speaker #1

  They suggest that these strategies might be capitalizing on behavioral biases in the market.

• Speaker #0

  We talked about those earlier. Hurting behavior, anchoring bias, the disposition affect all those human tendencies that can distort. market prices.

• Speaker #1

  Exactly. The idea is that these biases create predictable patterns in asset prices that can be exploited by systematic rule-based strategies.

• Speaker #0

  So it's almost like using algorithms to outsmart human irrationality.

• Speaker #1

  You could say that. Now, this is just a hypothesis, but it's supported by a growing body of research in behavioral finance, which is fascinating in its own right.

• Speaker #0

  It makes sense to me. I've seen firsthand how emotions can drive markets, sometimes to irrational extremes.

• Speaker #1

  We all have. And this paper provides further evidence that those emotions can create opportunities for algo traders who can design strategies that exploit these predictable patterns.

• Speaker #0

  That's food for thought. We've covered a lot of ground in this episode, and I'm really starting to grasp the power of these combined strategies. What are the key takeaways our listeners should keep in mind as they apply these ideas to their own algo trading?

• Speaker #1

  Let's distill it down to a few key points. First and foremost. This paper demonstrates the power of combining different strategies in algo trading. It's not just about finding one magic formula. It's about building a diversified and robust portfolio that can navigate a variety of market conditions.

• Speaker #0

  That's a great point. It's about the synergy between the strategies, not just their individual strengths. Right.

• Speaker #1

  And the specific strategies they highlighted, risk parity, momentum, and trend following each, play a crucial role in this combined approach.

• Speaker #0

  We saw how risk parity can help manage volatility. how momentum can help identify those high-performing assets, and how trend following can help us ride the wave but know when to get out before it crashes.

• Speaker #1

  Precisely. And the paper provides compelling backtest results to support the claim that this combined approach can lead to higher risk-adjusted returns and lower drawdowns, which is what we all want as traders.

• Speaker #0

  Those backtests were really impressive. They really highlighted the potential benefits of this approach.

• Speaker #1

  They did. And, perhaps most importantly, The paper reminds us that these strategies aren't just for those with PhDs in quantitative finance. They can be implemented by anyone who's willing to put in the time and effort to understand the concepts and build the necessary systems.

• Speaker #0

  So it's a message of empowerment. You don't need to be a Wall Street wizard to benefit from algo trading.

• Speaker #1

  Exactly. And with the increasing availability of tools and resources, it's becoming easier than ever to get started. There are platforms, software and educational materials that can help you build and back. test your own strategies.

• Speaker #0

  I think that's a great place to wrap up this episode. It's been an incredibly insightful deep dive, and I feel like I've leveled up my understanding of these powerful strategies.

• Speaker #1

  You too. It's always exciting to explore new research and discover new ways to approach the market.

• 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 backtest, Find us at https.paperswithbacktest.com. Happy trading.

• Speaker #1

  Yeah, it really highlights how even seasoned investors can be influenced by their own biases leading to suboptimal decisions.

• Speaker #0

  So you're saying that these seemingly complex strategies might actually be tapping into something quite fundamental. Right. Our own human nature.

• Speaker #1

  That's the fascinating part. This paper provides compelling evidence that these strategies risk parity, momentum, and trend following, especially when combined might be capitalizing on the very patterns created by those human biases.

• Speaker #0

  Okay, that's a mind-blowing concept. Yeah. So instead of fighting against these inherent biases, we can potentially use them to our advantage.

• Speaker #1

  Precisely. By understanding how these biases influence market behavior, we can develop algorithms that systematically exploit these predictable patterns.

• Speaker #0

  That's a powerful idea. And it seems like these strategies are capturing something unique, something that traditional risk models haven't fully accounted for.

• Speaker #1

  That's what the data suggests. The researchers went to great lengths to test various risk factors, and they found that a significant portion of the returns achieved by these strategies couldn't be explained by those conventional models.

• Speaker #0

  So it's not just about market risk or value or size. Right. There's something else at play here. Right.

• Speaker #1

  And the paper hypothesizes that this something else could be the very behavioral biases we've been discussing.

• Speaker #0

  It's like we're using data and algorithms to outsmart human psychology.

• Speaker #1

  In a way, yes. And this concept isn't just theoretical. It's supported by a growing body of research in behavioral finance, which is really transforming how we understand financial markets.

• Speaker #0

  It's incredible to think that these seemingly complex strategies might actually be rooted in something so fundamental the way our brains are wired.

• Speaker #1

  It's a reminder that even in the world of algo trading, human behavior still plays a critical role. And those who can understand and anticipate these patterns stand to gain a significant edge.

• Speaker #0

  This has been an incredibly insightful deep dive into the world of algo trading and behavioral finance. I feel like I've learned so much about not only how these strategies work, but also why they might be so effective.

• Speaker #1

  I agree. It's always exciting to explore the intersection of data, algorithms, and human psychology. It's a constantly evolving field with endless possibilities.

• Speaker #0

  Before we wrap up, can you summarize the key takeaways for our listeners, the things they should keep in mind as they design and implement their own algo trading strategies?

• Speaker #1

  Absolutely. First and foremost, this paper highlights the power of combining different strategies, risk parity, momentum, and trend following to create a more robust and potentially more profitable portfolio. It's not about picking one over the other. It's about finding the right balance that works for you.

• Speaker #0

  That's a great point. It's about synergy and diversification, not just chasing the hottest strategy at the moment.

• Speaker #1

  Right. Second, we need to remember that these strategies aren't just for the quants. Anyone can learn and implement them with the right resources and dedication.

• Speaker #0

  That's a message of empowerment. I'll Go Trading isn't some exclusive club. It's accessible to anyone who's willing to put in the effort.

• Speaker #1

  Exactly. And finally, this paper gives us a glimpse into the potential of behavioral finance to inform and improve our trading strategies. By understanding the patterns created by human biases, we can develop algorithms that exploit those patterns systematically.

• Speaker #0

  It's about using data and algorithms to gain an edge in a market that's still heavily influenced by human emotion. I think that's a powerful takeaway for all of our listeners.

• Speaker #1

  I agree. And it's a reminder that there's always more to learn and explore in the world of algo trading.

• Speaker #0

  This has been an incredibly insightful episode. I want to thank you for sharing your expertise and insights with us today.

• Speaker #1

  It was my pleasure. Always happy to dive into these fascinating research papers.

• Speaker #2

  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.