Focusing on Crisis Protection

---

This summary is written by Rich based on a conversation in our CTA series between Matt Dorsten, a Portfolio Manager within Quantitative Strategies at PIMCO, and the podcast hosts, Niels and Alan.

About PIMCO

Niels and Alan were joined by Matt Dorsten, a Portfolio Manager within Quantitative Strategies at PIMCO. 

They discussed PIMCO's systematic approach to bond investing and their entry into direct systematic investing with tail risk hedging 15 years ago, followed by trend following. 

PIMCO currently manages about $60 billion in assets under their quant investing umbrella, with trend following being the second largest segment at around $5 billion. 

Matt joined PIMCO around the time they started expanding into systematic investing. The conversation will focus on topics related to trend following and managed futures which comprise a part of PIMCO’s portfolio of Programs. 

Matt Dorsten discussed PIMCO's approach to trend following, which began with tail risk hedging and evolved to focus on defensive properties. They prioritize defensive characteristics over absolute returns when constructing their portfolio. 

Matt explained that trend following is unique as it can profit during crises and doesn't cost money to implement. The defensive characteristics persist because trend following allows shorting assets in a downturn, unlike other strategies, and behaves similarly to a long options position with a convexity effect in crisis events. 

The fact that it pays to implement this strategy is attributed to behavioural anomalies that many investors face. Behavioural anomalies refer to irrational decision-making or cognitive biases that can impact investors' performance. 

PIMCO prioritizes defensive characteristics over absolute returns when constructing their portfolio, meaning that they focus on protecting investments during market downturns and periods of increased volatility. This approach can provide valuable diversification benefits and help reduce the overall risk of an investment portfolio. Here are some key aspects of this strategy:

• Crisis protection: Defensive strategies aim to perform well or at least maintain value during market crises, when many other investments may experience significant losses. This can help preserve capital and cushion the impact of market declines on the overall portfolio.
• Low correlation: By prioritizing defensive characteristics, the portfolio seeks investments that have a low or negative correlation with traditional assets like equities. This can help reduce the overall portfolio risk and provide diversification benefits during different market environments.
• Risk management: Focusing on defensive characteristics means that risk management is a key consideration when constructing the portfolio. This can involve regularly monitoring and adjusting portfolio exposures, as well as using various risk management techniques to minimize potential losses.
• Long-term perspective: Prioritizing defensive characteristics often requires a long-term perspective, as the benefits of a defensive approach may not be immediately evident during periods of strong market performance. However, the long-term benefits of improved risk-adjusted returns and reduced portfolio volatility can be significant.
• Selective investment approach: With a focus on defensive characteristics, the portfolio construction process may involve a more selective approach to investments. This can result in a smaller number of carefully chosen positions that are expected to perform well in various market conditions.

By prioritizing defensive characteristics over absolute returns, PIMCO aims to create a trend-following strategy that provides valuable protection during market downturns and helps to reduce the overall risk of an investment portfolio. This approach sets them apart from other trend followers who may focus more on maximizing returns without necessarily considering the defensive aspects of their strategies.

Niels asked Matt if PIMCO's trend following strategy behaves differently compared to their peers due to their focus on defensive properties. Matt acknowledged that while the broad performance behaviour is similar, their strategy has the most negative correlation with equities, which confirms that their approach is working as intended.

Matt also mentioned that performance is important to them, and they aim to create a strategy that is both defensive and high returning. However, the primary driver behind their strategy remains its defensiveness. This perspective sets them apart from other trend followers who may put more weight on risk-adjusted metrics when designing their strategies.

The Challenge of Trend Following for Investors

Matt and Niels discussed the challenges of getting investors to view trend following in a long-term perspective, rather than as a trade. They noted that people often find it hard to stomach the seemingly random return profile, prolonged drawdowns, and the fact that it can have drawdowns during times when equities are performing well.

According to Matt, trend following's basic premise contradicts what most people think about investing, which makes it difficult for them to embrace. However, he emphasized that investors should expect diversifying strategies like trend following to have drawdowns at unusual times, as it is precisely this low correlation with other strategies that adds value to their portfolios. 

Although it is challenging to change investors' mindsets, educating them about the long-term benefits and diversification provided by trend following remains important.

Methods of Creating Defensive Portfolios

Alan asked Matt about the methods they use to build a more defensively oriented trend-following program. 

Matt reiterated that they avoid diluting the strategy's defensive characteristics by not adding diversifying strategies to the portfolio. They also employ a relatively fast trading speed, which, while not affecting the absolute expected return, makes the strategy more defensive during market drawdowns. However, this also results in higher turnover and false starts, requiring careful attention to execution costs.

Additionally, Matt's firm adopts an asymmetric approach to equity risk, not only in outright equities but also in assets correlated with equities, such as long oil, long base metals, and long emerging market currencies. 

By limiting potential exposure to these risk-on assets, the strategy becomes more defensive and mitigates the risk of being caught on the wrong side during a sudden sentiment reversal, like the one experienced in February 2018.

Alan asked Matt about the risks of being faster in their trend-following strategy, particularly during prolonged equity bear markets with counter-trend moves. 

Matt said that their average holding period is one to two months, which is faster than the industry average of around six months. However, he admited that there is no easy solution to protect against being shaken out of a trend during a bear market rally, as it's part of the territory. 

For example, they experienced this issue in the previous year when they exited a short position during a bear market rally, which turned out to be a bad move.

Niels asked Matt about balancing risk-on trades and letting profits run while maintaining a defensive position in their trend-following strategy. Matt explained that having a broad universe of assets helps them find non-equity-correlated trends that can be profitable. He cited the European energy crisis as an example of a large trend not correlated with equities. Matt also pointed out that there is no correlation between ex ante equity beta and subsequent realized return in trend following, which is unique compared to other assets.

When asked about using machine learning in trend following, Matt said they have used it in other contexts, such as equity strategies, but have not found any obvious uses in the trend-following space. He thinks machine learning in this context is more of a buzz topic and may not necessarily pay off in real life.

Alan asked Matt about the role of bonds in the portfolio and how it affects their equity beta calculation. Matt explained that bonds are factored into their equity beta calculation, but they don't make any discretionary calls. He also mentions that correlations take a long time to measure robustly. Matt believes that bonds still have the same behaviour they've always had in crisis events.

When asked about incorporating a long VIX or long volatility strategy as part of the defensive mix, Matt said they run that type of strategy in their other multi-strategy quant portfolios but not in their pure trend-following portfolio.

Regarding constructing a portfolio of equities and defensive strategies, Matt suggested using a mix of different strategies, as each of them comes with its own limitations or costs. He prefers trend following because it pays to diversify and provides protection over time through asset allocation.

Research and Enhancements

Matt discussed how the trend-following strategy has evolved over time, with gradual changes in the basic strategy design and significant developments in expanding their universe. He emphasized the importance of diversification, as most trades are losers, and capturing rare profitable trends is essential.

Niels asked if the type of trend-following methodology used makes a significant difference. 

Matt believes it makes very little difference, and the focus should be on the speed of the ultimate signal instead. 

In terms of trading more markets, Matt argued that the performance of exotic markets has been more consistent historically, with less degradation of the Sharpe ratio compared to more commonly traded markets. 

He suggested that trend-following performance has degraded in more liquid and commonly traded markets, while exotic markets have experienced a smaller decline.

Fees and Trend Replication

Matt responded to the question of whether trend-following strategies are easy and whether they should be low-cost. 

He argued that while the principle may seem simple, the details and execution are complex. Trading a large number of markets, especially exotic ones, requires substantial effort and expertise to manage the data and minimize transaction costs. 

Matt also addressed the topic of trend replication through linear regressions on fund returns. He believes that such an approach would not capture the dynamic nature of positions, especially at critical turning points, and would likely miss unusual trends in markets like Hungarian interest rates or European energy. Therefore, he does not think trend replication through linear regressions is an effective method.

Has the Effectiveness of Trend Trading Degraded Over Time?

Alan asked Matt about the effectiveness of fast trend trading and whether it has degraded over time. 

Matt acknowledged that there were periods when fast trend trading was more successful, but he doesn't see a robust conclusion that returns have significantly degraded. He believes that the data points more to random noise and market-specific events, rather than a structural change.

Matt also shared his perspective on trading alternative markets and how trading costs in these markets influence their model application. They tend to trade slower in less liquid markets and faster in more liquid ones, maintaining a uniform approach to avoid data mining issues.

Regarding very fast trading systems, Matt explained that while they see value in them, they are not very scalable. The focus is on constructing a portfolio that is capacity efficient and allows for a diverse universe of markets.

Lastly, Matt addressed two questions from a large allocator. He agreed with capping equities but not to an extent that it significantly reduces return potential. Regarding non-price-based trend signals, he supported adding them if they exhibit crisis alpha. However, he wouldn't add them if they degrade portfolio defensiveness.

On Diversification

Niels asked Matt about the number of trend following strategies needed in a portfolio. Matt suggested having a small set of diverse strategies to maintain a balanced portfolio, as adding too many can result in a more normally distributed return profile. 

As for the second diversification option after trend following for a traditional 60/40 portfolio, Matt mentioned duration, long Japanese yen, and gold as possibilities, but he believes trend following is more reliable. 

When the topic of long volatility was brought up with respect to diversification benefits, Matt expressed his opinion that it's too expensive to be a practical diversification option in a general context. Long volatility strategies aim to profit from an increase in market volatility, but they can be costly to maintain, and their performance might not consistently align with investors' expectations.

Framing Performance Outcomes to Investors

Alan asked Matt how they frame expected outcomes for their investors and what metric they use to measure the effectiveness of their trend following program in delivering its defensive quality. 

Matt explained that they aim to maximize the Sharpe ratio conditional on being in equity market drawdowns. This means they focus on the performance of the strategy during unfavourable market conditions. However, they still work to increase the absolute Sharpe ratio, for instance, by having a broad investment universe.

In terms of describing the program to investors, Matt said that they expect it to perform well during significant equity market drawdowns. They often analyse quintiles of equity quarterly returns and target good performance in the worst quintile, which represents the most extreme market situations. 

They expect the strategy to be positive on average during these large drawdowns.

Liquidity

Alan asked Matt about liquidity, particularly given that they are a faster trend follower than average, and whether there have been any notable shifts in liquidity conditions over time. 

Matt responded that their focus is solely on liquid derivatives. He believes the changes in market conditions have impacted cash securities more, as those require a balance sheet to intermediate risk. Matt doesn't think they've seen any significant change in liquidity in the assets they trade.

Niels jokingly asked Matt if he plans to publish a book called "The Trend King," referencing a book about Bill Gross, a former employee of their firm, titled "The Bond King." Matt modestly responded that he would never put himself in the same category as Bill Gross, who is a towering figure in the industry.

Misconceptions of Trend Following

Niels asked Matt about the one thing he hears about trend following that he disagrees with the most. 

Matt shared that he often sees headlines about trend followers being long or short equities and how that might affect the S&P 500 or 10-year bond futures. He disagreed with this notion, stating that trend followers are too small to have an impact on those liquid markets, and he considers such headlines to be more like clickbait. 

Outlook for 2023

Matt discussed his outlook for 2023 and the factors that may impact the trend following landscape. 

He is encouraged by the reduced presence of central banks in global markets. Matt recalled a period in the past when trends were disrupted by active, globally coordinated central banks, and he believes that the current situation, with central banks facing multiple challenges due to post-COVID inflation, could create a more favourable environment for trend following. 

He also noted that the Bank of Japan might potentially remove some of its extraordinary measures, leading to more trends forming as market themes are less controlled and freer to move.

Matt believes that being part of a bond-focused organization like PIMCO has been beneficial for their trend following strategies. 

Although the value-oriented investors at PIMCO provide a healthy dose of skepticism, the collaboration with discretionary traders has been helpful in exploring new market opportunities. 

Matt emphasized the importance of having material exposure to a broad universe of instruments for success in trend following. Being part of a large organization like PIMCO has provided them access to data, expertise, and a wide range of markets that contribute positively to their approach.

---

This is based on an episode of Top Traders Unplugged, a bi-weekly podcast with the most interesting and experienced investors, economists, traders and thought leaders in the world. Sign up to our Newsletter or Subscribe on your preferred podcast platform so that you don’t miss out on future episodes.