Tracking major macro risks
Earlier this year, investors across the world were trying to evaluate the impacts of sticky inflation and rising rates on their investments—and we were no exception.
Our global economic outlook showed a perfect storm of headwinds ahead. Central banks had pulled the parachute on economic growth and, as a result, business and consumer sentiment were softening. As an institutional investor, we needed a disciplined process to track our risk exposure as markets evolved.
Our Global Multi-Asset Team’s house view provided a valuable baseline for our approach, which kept major risk drivers on our radar. But our Disciplined Equity team wanted to go a step further and build a targeted analytical framework around interest rate sensitivity. So, as an overlay to the macro lens, we constructed a machine learning model to monitor the knock-on effects of rising interest rates. (For a closer look at Multi-Asset Team’s house view, see Appendix 1 below.)
How our MLSIR model works
Yield curves have inherently represented investors’ expectations across a range of macroeconomic factors. For example, markets consider the inversion of Treasury yield curves as a predictor of recession because interest rates encode beliefs about the broader environment. By isolating those signals in the noise, we believe we can learn how to improve our portfolio construction.
Our proprietary Machine Learning Sensitivity to Interest Rate model (MLSIR) isolates recession fears, inflation concerns and interest rate expectations. By leveraging this system within our investment process to manage risks we believe we can deliver a more consistent outperformance in equity markets over complete market cycles.
One reason we built the model was because of weaknesses in the traditional Optimal Least Squares (OLS) method. OLS estimators are capable of processing tons of data through standard regression, but they cannot always determine whether the results are statistically significant or not. In other words, we do not have a high degree of confidence on how the output generated from the model should be weighed in our decision-making process.
Yield curves inherently represent investors’ expectations across a range of macroeconomic factors. Why else do markets consider the inversion of Treasury yield curves as a predictor of recession?
By contrast, our MSLIR model uses an elastic net regularization method to estimate which findings are statistically robust, and which can be zeroed out. The output affords us a ground up view on each stock’s individual sensitivity—not only to interest rates, but to recession fears and inflation as well.
One edge of our MLSIR model is to distill the pure breakeven inflation rate. Typically, this is calculated by taking today’s interest rate level minus the inflation-linked bond level. However, within that component, how much can we explain, and how much is noise?
In our view, a component of the breakeven inflation will be impacted by the Fed’s future decisions. As rates go up, that component will come down. But the second piece of breakeven inflation, which we refer to as pure breakeven, will not be affected by monetary policy actions. It could be caused by geopolitical risk or supply chain disruptions—but regardless, these risks must be accounted for in order to truly protect capital over the long term.
Layering it over the portfolio
We currently implement the MSLIR model in the BMO Canadian Smart Alpha Strategy, which tilts towards low volatility equities. Combining low vol’s natural defensive qualities with our risk overlay allows the portfolio to be consistently less inflation-sensitive than the broad benchmark. It is also well-positioned for a recession, offering better capital preservation.
Positioned for higher recession risk
Bloomberg, BMO Asset Management Inc to July 31, 2023.
MLSIR macro betas by sector
Data date: March 31, 2023. Source: Bloomberg, BMO Asset Management Inc
We apply the MSLIR model through each stock in the investment universe to determine whether they have significant sensitivities or not. As mentioned, any old regression program would churn out a coefficient for the relationship; ours goes a step further to tell us if that beta value is statistically meaningful. We also run historical simulations in different market periods and with anomalous data to stress test the integrity of our findings.
Crucially, insights from the MLSIR contribute to risk management and portfolio construction at the input phase. We exercise it as an analytical tool, in keeping with our broader use of quantitative strategies. We are not hands-off programmers letting the portfolio manage itself—we are disciplined managers who utilize machine learning for systematic execution of our strategies. Consider the parallel of autonomous vehicles: even if on-board sensors can help with assisted driving, we need to be in the car at all times with both hands on the wheel.
The evolution of quant modeling
Over the past 15 years, I have seen numerous approaches to quantitative investing while working for some of Canada’s largest institutional asset managers. I have actually written an equity model in APL, the oldest programming language in finance. It was originally created in the 1960s and uses Latin syntax that reads from right to left. Eventually, when the firm needed to update its code, I helped translate these fundamental models into SAS.
Years later, I witnessed one of the first AI algorithms being implemented at a major pension fund. During that era, I also transitioned into a more senior role focused on equity investments. Undergoing these experiences simultaneously only strengthened my belief that true disciplined investing can be accomplished through a combination of quantitative and traditional investment valuation methods.
At BMO GAM’s Disciplined Equity team, I found like-minded professionals who believed in a “human plus machine” version of quantitative investing. Our team has diverse views and a variety of different experiences, which we use to inform a culture of vigorous debate and discussion. Some members of the team hail from a quantitative background like myself while others specialize in asset allocation and research, and we also have a dedicated team of three investment engineers focusing on data architecture.
The depth and diversity of BMO GAM’s talent is central to how we seek to optimize risk-adjusted returns. As with the MLSIR model, having such a broad range of expertise helps us identify and develop the tools that institutions need to successfully navigate complete market cycles. With central banks continuing to hike interest rates undeterred by rising inflation and recession risks, machine learning is one such tool at our disposal to help manage risks and ultimately protect clients’ capital.
Please contact your BMO Institutional Sales Partner to learn more about the Disciplined Equity team, the MLSIR model and the BMO Canadian Smart Alpha Strategy.
Disclosures
Not intended for distribution outside of Canada.
Certain of the products and services offered under the brand name BMO Global Asset Management are designed specifically for various categories of investors in a number of different countries and regions and may not be available to all investors. Products and services are only offered to such investors in those countries and regions in accordance with applicable laws and regulations.
This communication is intended for informational purposes only and is not, and should not be construed as, investment, legal or tax advice to any individual. Particular investments and/or trading strategies should be evaluated relative to each individual’s circumstances. Individuals should seek the advice of professionals, as appropriate, regarding any particular investment. Past performance does not guarantee future results.
Any statement that necessarily depends on future events may be a forward-looking statement. Forward-looking statements are not guarantees of performance. They involve risks, uncertainties and assumptions. Although such statements are based on assumptions that are believed to be reasonable, there can be no assurance that actual results will not differ materially from expectations. Investors are cautioned not to rely unduly on any forward-looking statements. In connection with any forward-looking statements, investors should carefully consider the areas of risk described in the most recent simplified prospectus.
Commissions, trailing commissions (if applicable), management fees and expenses all may be associated with mutual fund investments. Please read the fund facts or prospectus before investing. BMO Money Market Fund is not covered by the Canada Deposit Insurance Corporation or by any other government deposit insurer. There can be no assurances that the Fund will be able to maintain its net asset value per security at a constant amount or that the full amount of your investment in the fund will be returned to you. Past performance may not be repeated.
BMO Mutual Funds are managed by BMO Investments Inc., which is an investment fund manager and a separate legal entity from Bank of Montreal.
BMO Global Asset Management is a brand name under which BMO Asset Management Inc. and BMO Investments Inc. operate.
®/™ Registered trademarks/trademark of Bank of Montreal, used under licence.
