Transcription

• Speaker #0

  Hello, welcome back to Papers with Backtest podcast. Today, we dive into another algo trading research paper. We're cracking open betting against beta. Ah, right. Which has a title that just screams contrarian strategy.

• Speaker #1

  It does. Yeah, it grabs your attention, doesn't it? But this paper by Andrea Frazzini and Lassa Haida-Peterson is more than just a catchy title. It really challenges kind of a fundamental assumption we have in finance, which is that higher risk equals higher reward.

• Speaker #0

  OK, let's unpack that a bit. You know, we talk about the risk return tradeoff. Every investor, you know, hears about it. What's the twist here? Well,

• Speaker #1

  the twist is leverage or more precisely, I guess, the constraints on leverage that most investors face. Think about it this way. You have two assets and they both have the potential to reach a certain return target. One's a bit safer, but you need a ton of leverage to hit that target. The other one is, you know, inherently riskier, but needs less leverage for that same return. So if you're an investor that can't easily access, you know, mountains of leverage, you might be forced into that riskier asset, even if it doesn't actually offer, you know, proportionally better returns for the risk you're taking on.

• Speaker #0

  So you're stuck in the slow lane, even if you're willing to floor it.

• Speaker #1

  Exactly. Yeah. And this paper argues that that actually happens systematically in the market. So it kind of throws that whole traditional risk reward relationship out of whack, which creates opportunities for those who can spot that mispricing.

• Speaker #0

  And that's where this betting against beta comes in.

• Speaker #1

  Exactly. Yeah. So the paper proposes a strategy along for a portfolio, which they call the B-AB factor.

• Speaker #0

  OK, break that down. The B-AB factor. How does it actually work?

• Speaker #1

  Sure. So what they do is they start by identifying a basket of low beta assets. So these are assets that are, you know, expected to be less volatile than the overall market. Right. And then they leverage up this basket to achieve a beta of one.

• Speaker #0

  OK, hold on. A beta of one. Just remind us quickly what that means again.

• Speaker #1

  Sure. So beta is essentially just a measure of an asset's volatility relative to the market. So a beta of one means it tends to move, you know, kind of in line with the market up or down. So they're essentially amplifying the movements of these, you know, typically less volatile assets to match those market swings.

• Speaker #0

  Got it. So bringing those low beta assets up to like a little playing field, right? What about the other side of that?

• Speaker #1

  Yeah. So on the flip side, they identify a basket of high beta assets. So these are the ones that are expected to be more volatile, right? And then they deleverage this basket down to a beta of one. So they're basically creating a level playing field in terms of risk, adjusting for those inherent volatility differences between the low beta and high beta assets.

• Speaker #0

  So they're long. Those potentially undervalued, less volatile assets leveraged up and short those potentially overvalued, more volatile assets deleveraged down. All with that beta of one. What kind of results did they actually see when they tested this out?

• Speaker #1

  Well, here's where it gets really interesting, especially for, you know, algo traders like us. They backtested this BIB strategy across various asset classes, different time periods. And, yeah, they found statistically significant positive returns.

• Speaker #0

  Yeah. Positive returns always sound good, but we know risk-adjusted returns are really where it's at. So what about that? What did they find?

• Speaker #1

  Right. Yeah. So they calculated the Sharpe-Ray ratio, which, you know, gives you a sense of the return per unit of risk taken. And the B-I-B strategy consistently showed a really, really high Sharper Rate Ratio, which suggests, you know, a potentially much more efficient way to manage your risk and allocate your capital.

• Speaker #0

  Okay, so that's a pretty compelling argument, I think, at least to consider this, you know, betting against beta approach. But before we go too far, let's dive into the specifics of, you know, their trading rules and how those backtests actually performed. What were, like, the key parameters they used?

• Speaker #1

  Yeah, absolutely. Let's get into the nuts and bolts of their actual trading rules. So they focused on a monthly rebalancing approach.

• Speaker #0

  Monthly. Okay. So not like some crazy high frequency signal chasing system, right? Yeah. This is more about capturing kind of broader market inefficiency.

• Speaker #1

  Exactly. Yeah. So each month they would rank assets based on their betas calculated over, you know, the previous year. And then they would divide those assets into deciles. So basically the, you know, the ones with the lowest betas in one group, the ones with the highest in another.

• Speaker #0

  OK, so you're creating those two kind of distinct baskets, right? One for going along, one for shorting and updating them each month based on how how volatile those assets have been.

• Speaker #1

  Exactly. Yep. And then and then they would construct their long short portfolio by longing that lowest beta decile, leveraging it up to a beta of one and then shorting the highest beta decile. deleveraging that down to a beta of one.

• Speaker #0

  So it's almost like a systematic contrarian approach in a way, right? Buying the boring stuff, betting against the exciting stuff, but all while controlling for risk by managing that beta.

• Speaker #1

  You got it. Yeah. And this is where their back tests get really, really interesting. So they tested this strategy across a range of asset classes, and the results were pretty compelling. So for U.S. stocks, they found an average monthly return of around 1% above the market over a period of several decades.

• Speaker #0

  Wow. One percent a month. That's that's significant, especially, I mean, considering it's coming from like a relatively simple rules based approach. Right. But did those returns hold up in other markets?

• Speaker #1

  That's the remarkable part. Yeah. They found similar patterns, not just in U.S. stocks, but also in international markets and bonds, even in commodities. OK.

• Speaker #0

  Now, I mean, it's important to note, right, no strategy works perfectly in every single market environment. Were there any do they highlight any specific periods where this. BAB strategy struggled?

• Speaker #1

  That's a really insightful question. And it gets to, I think, the heart of how to apply this research in a practical way. So one thing they found was that the BAB strategy tended to underperform during periods when you had what they call tightening funding constraints. Okay.

• Speaker #0

  Funding constraints. Give us a real world example of what that might look like.

• Speaker #1

  Sure. Imagine a sudden market shock or like a financial crisis. Investors get spooked. Lenders get more risk averse and suddenly accessing leverage becomes much harder and more expensive for everyone.

• Speaker #0

  Oh, so it's like musical chairs when the music stops. Everyone's scrambling for a seat.

• Speaker #1

  Exactly. Yeah, yeah. And in those moments of kind of panic, the authors found that the BAB factor, which relies on exploiting a certain market dynamic, just it didn't perform as well.

• Speaker #0

  OK, so that's crucial, I think, for an algo trader to understand, right? You can't just blindly program in these rules and expect it to. to work flawlessly in every market condition. You really need to be aware of those particular pitfalls and adjust accordingly.

• Speaker #1

  Absolutely. Yeah. I think it just reinforces the importance of, you know, risk management and understanding the limitations of any strategy, really. But here's the silver lining. They also found that the BAB factor tended to bounce back really strongly once those periods of extreme funding constraints eased.

• Speaker #0

  Interesting. Okay. So it's not about necessarily, you know, timing the market and avoiding those periods altogether, but more about, like you said, managing your risk effectively through those periods. Right. And then potentially even capitalizing on the rebound when things start to loosen up again.

• Speaker #1

  Yeah, absolutely.

• Speaker #0

  What other, were there any other kind of key takeaways that they highlighted for traders who might be looking to, you know, actually implement a strategy based on these findings?

• Speaker #1

  Yeah. So one crucial insight, I think, was the importance of transaction costs. So their analysis showed that even small differences in trading costs could significantly impact the performance of the BAB strategy.

• Speaker #0

  Makes sense, right? Especially with a long, short approach, you're involving leverage. You know, even those seemingly insignificant fees can really kind of eat into your returns over time. So any algo you build around this would need to factor that in very carefully.

• Speaker #1

  For sure.

• Speaker #0

  What about, you know, thinking about the specific assets used in the baskets, right? Did they find that certain types of stocks or bonds, for example, performed consistently better than others within this framework?

• Speaker #1

  That's a great question. And the paper actually goes into quite a bit of detail on that. They break down their findings across different sectors, industries, even individual stock characteristics.

• Speaker #0

  This is getting really interesting. Let's dive into those specifics in our next segment, because I know our listeners who are all about that practical application, they're eager to hear how those nuances kind of play out. And building a real world trading algorithm. Okay. So let's get granular then. We're talking about, you know, specific assets. Did they find any sectors, any industries that kind of consistently popped up in those high performing low beta baskets?

• Speaker #1

  Yeah, they did. So in the U.S. stock market, they found that sectors like utilities, consumer staples, you know, things like essential goods and services, those tended to fall into that low beta category pretty often.

• Speaker #0

  Interesting. Yeah, these are. These are sectors that are often, you know, considered kind of less volatile, I guess. Right. Less sensitive to those economic ups and downs.

• Speaker #1

  Exactly. Yeah. And because they're not as, you know, sexy, I guess, as like tech stocks or something, they might be overlooked by investors who are, you know, seeking those really high-flying returns.

• Speaker #0

  Yeah. Which could then create that mispricing opportunity that this whole BAB factor is trying to exploit, right? Exactly. Now for... algo traders, this is like gold, right? We're talking about potentially predictable characteristics that you can then use to try to identify those undervalued assets. Did they find similar patterns in other markets too?

• Speaker #1

  They did, yeah. Although, you know, the specifics varied a bit by region, by asset class. But for example, in international markets, they found that more mature dividend paying companies tended to fall into those low beta buckets more often.

• Speaker #0

  So again, a little bit of a contrarian signal there. Instead of chasing high growth, high risk companies, this approach might lead you to those kind of steady dividend paying stalwarts that are Maybe less exciting in terms of the news, but potentially, you know, significantly undervalued.

• Speaker #1

  Yeah, exactly. And that's what I think makes this research so valuable, especially for, you know, for algo traders. It gives you a framework, a set of rules, essentially, that you can test and refine based on, you know, your specific goals, your risk tolerance.

• Speaker #0

  And importantly, right, they've provided those concrete backtest results so you can actually see, you know, not just if it works, but how well it's worked historically. Right. Were there any other... Any other kind of key insights from their back test that you think would be particularly useful for our, you know, algo-minded listeners?

• Speaker #1

  Yeah. One thing they highlighted was the importance of, you know, portfolio construction. So it wasn't just enough to, you know, sort assets into these low and high beta buckets. They found that incorporating additional factors like momentum or value could really further enhance returns.

• Speaker #0

  So it's not just about the beta. It's about combining it with these other signals to create kind of a... A more robust strategy, I guess, like adding spices to a recipe to really bring out those complex flavors.

• Speaker #1

  Yeah, I like that analogy.

• Speaker #0

  Do they find like a particular, you know, magic combination of factors that seem to work best?

• Speaker #1

  You know, they explored various combinations and there wasn't like one, you know, magic formula. But their research suggested that that blending beta with. momentum signals in particular showed some really, really promising results. But again, I think this is where, you know, your own backtesting and optimization come in, right?

• Speaker #0

  Of course, yeah.

• Speaker #1

  What works best historically might need to be, you know, tweaked and adjusted based on current market conditions or your own, you know, individual risk appetite.

• Speaker #0

  Absolutely. And I think that's a perfect note to kind of end on, right? This paper, Getting Against Beta, it offers... you know, I think a really fascinating challenge to those traditional assumptions that we have about risk and reward. It provides this practical framework for potentially, you know, profiting from these market inefficiencies. And I think it highlights, you know, for me, really highlights the importance of combining multiple signals, right? Managing your risk effectively and always being willing to adapt to those changing market dynamics that are always, you know, throwing us curveballs.

• Speaker #1

  Well said. Yeah, I think it's a great reminder that sometimes the most profitable opportunities are kind of hiding in plain sight. They're in those less crowded corners of the market, just waiting to be discovered by those who are willing to, you know, to bet against the herd.

• Speaker #0

  Love it. Well, thank you so much for diving into this paper with me. It's been a fascinating discussion.

• Speaker #1

  My pleasure.

• Speaker #0

  And to our listeners, thank you for tuning in to Papers with Backtest podcast. We hope this deep dive gave you some useful insights. Join us next time as we break down more research and figure out how to turn it into practice. And for more papers and backtests, find us at paperswithbacktest.com. Happy trading!