Systematic fixed income explained | Insight Investment

Systematic fixed income investing has finally come of age.

It has been a long road to get here. Our team has been there from the start developing systematic credit strategies for over 20 years.

What is systematic investing?

Well, it's a quantitative rules-based approach to managing assets in corporate bonds. For example, we don't invest based on credit analyst recommendations or portfolio manager's view. Instead, we build quantitative models to assess bonds along multiple risk and return dimensions, and then construct portfolios that aim to exploit market inefficiencies.

To understand the origins of systematic investing, I can go back to the late 1960s when a man named John McQueen assembled a dream team of data-driven revolutionaries or quants here in San Francisco to challenge the traditional ways of investing in stocks.

In the 1970s, bill faus and Tom Loeb spun out of this team to design and run the world's first equity indexing strategy. This was unique at the time because it prioritized quantitative investment methods instead of supposed stock picking skills.

Fast forward to the 1990s when systematic investing took another major step forward. When University of Chicago economists, Eugene Fama and Kenneth French published their famous three factor model, this really kicked off the birth of factor-based investing in the equity world.

Now, whereas systematic equity strategies are so common today that even a bedroom coder can participate until very recently, fixed income investment approaches remains somewhat stuck in the old school traditional world.

Why? The main factor was lack of liquidity. Many bonds are difficult and costly to trade. It took the rise of fixed income ETFs in the two thousands to really change the game.

Our teams deep experience in the ETF world enabled us to pioneer credit portfolio trading, a trading protocol that can meaningfully reduce transaction costs and enhance liquidity in some markets.

Another key development was the proliferation of quality bond market data. This allows us to create increasingly powerful models that can aim to harvest alpha in very different ways than the traditional stock picker, or in this case, bond picker.

For example, we can build models to calculate what the fair price of a bond should be based on underlying observable market fundamentals, which we can then compare to traded market prices to identify which bonds might be expensive and which bonds might be cheap.

With decade long track records in systematic high yield credit, we can now confidently say that we believe our approach has the potential to produce robust outcomes that not only may produce consistent alpha, but are also very different from traditional discretionary strategies.

Our high yield alpha strategy, for example, shows a negative alpha correlation with the top high yield managers. Remembering the adage that diversification is the only free lunch. This feature alone can be critical to a multi-manager approach.

Therefore, in addition to potentially offering better returns with greater consistency and enhanced liquidity, our clients often see systematic investing as a diversifying compliment to their discretionary investment portfolios.

If you would like to learn more or have any questions about systematic investing, please contact us.

We think most market inefficiencies often have systematic solutions.

The objective of a systematic application is to exploit market inefficiencies by applying structured rules-based quantitative insights.

Systematic or quantitative techniques have been applied to high risk areas like stocks for over five decades with strategies targeting everything from value, volatility, momentum, yield, and beyond.

In fixed income, systematic approaches are actually not suitable across every asset class. In areas where liquidity is plentiful and dispersion is low, like treasuries, mortgages, and agencies. In our experience, systematic approaches are less effective and probably not needed. These asset classes are, to put it bluntly, too efficient, but in other areas like credit, the story is very different.

We believe systematic approaches can be particularly successful in high yield and also investment grade. Our research suggests they can also work in emerging market bonds. These markets are typically less efficient for multiple reasons.

We'll focus on high yield in another video. Here I'll focus on local currency, emerging market bonds.

Local currency, em bond investors often need to contend with a combination of headwinds like withholdings, taxes, moderately high trading costs, and operating hurdles like restricted currencies and sub custodian agreements.

We estimate that they can lead to a performance drag of anywhere between 25 and 50 basis points per year. This means that a purely passive investment approach is likely to lead to what we call an index minus outcome.

A traditional active strategy might be one way to overcome this, but our research indicates that a drawback to these strategies is that they tend to be relatively concentrated in our view.

A systematic approach here can aim to deliver an elegant solution by investing in a highly diversified portfolio using models to implement positioning.

An example of one of our models is the term premium signal. This aims to identify attractive country opportunities where a country's bond yield is higher than its cash yield. Buying bonds offer a yield pickup.

This signal allows us to systematically analyze every country and determine the most potentially compelling yield pickup and seek higher return potential.

If you have any questions about applying systematic strategies to fixed income, please do not hesitate to contact us.

Thank you.

Developing fixed income Alpha models requires specialized experience and expertise. But what makes a successful Alpha model?

At Insight, we believe that any strategy needs a combination of alpha drivers, that together meet three key qualities.

Number one, is there a clear economic rationale? Even if something's worked in the past, we need an economic reason to believe it will work in the future.

Number two, does it perform even if it's economically sound? Do we expect it to deliver outperformance on a structural and reliable basis going forward?

And number three, does one complement the strategies? Other alpha drivers?

In our view, ideas with the high correlations are of limited use in high yield credit markets.

We've spent over a decade implementing three alpha models together, quality, value, and structural themes.

The first one, quality focuses on downside protection by flagging companies with potentially deteriorating fundamentals and even risk of default. This offers a clear economic and performance justification. We expect quality to work well in down markets.

However, there are market conditions where it might not perform chiefly when markets are rising, which is when weaker credits have often done well. Therefore, we would not deploy quality in isolation.

The second is our value model. It looks to identify mispriced bonds or those where the market price is significantly different to the price we think is implied by credit fundamentals alone.

In our experience, this model has worked best through periods of high dispersion and volatility, checking both the economic and performance boxes. Crucially, it's tended to perform well, not just in down markets, but also in up markets, thus offering diversification to our quality model.

The third model structural themes seeks to exploit market inefficiencies like exploiting predictable volatility around new issues or downgrades from investment grade to high yield, otherwise known as fallen angels. Exploiting market inefficiencies offers clear economic and performance justifications to us.

Historically, we find it's also worked well through benign, less volatile market regimes, which has been the one area in which the other two models quality and value have been less likely to outperform.

The alpha models we apply may be different elsewhere, like an emerging market debt. Here we've carefully curated two different alpha models. We call them our term premium and currency signals.

The term premium signal seeks to identify country specific risks to identify likely winners and losers and countries where risks may be uncompensated by the returns.

The FX signal seeks alpha from emerging market currency positioning by identifying macroeconomic factors such as growth inflation, and indicators of stress.

The two models have as designed historically, have had a negative correlation with each other, checking all three of our boxes.

Insights team has been researching and developing systematic alpha drivers for over 30 years.

If you're interested in a deeper look into our research modeling and strategy construction and how it may add value to your portfolio, please do not hesitate to contact us.

Thank you.