How to leverage technology to generate alpha in an inverted yield curve environment ?
Here we are : August, 14th 2019, the yield curve inverted. Maybe the signal of a future recession but certainly the best example of the biggest challenge for Fixed Income Managers. For them, the struggle will become harder and harder. And it is not getting any better in the years to come.
“ Investors are spooked by a scenario known as the “inverted yield curve,” which occurs when the interest rates on short-term bonds are higher than the interest rates paid by long-term bonds. What it means is that people are so worried about the near-term future that they are piling into safer long-term investments.” Jonnelle Marte said in her article in the Washington Post.
For Fixed Income managers, it is getting harder to generate alpha. At the same time, more and more corporate bonds carry a negative yield. With this, the risk of a setback in the market is increasing from day to day.
For many investors in the Asset Management industry, this is a huge concern. The dilemma is about holding or selling a position, or taking more risk by going into high yield position…
From my point of view, this question should be : Isn’t this the right moment to reshape the way we invest? Should we use this tense period in financial markets as an opportunity to reshape the investment process and leverage technologies in our decision making? Can we use algorithms to enhance portfolio management ?
Hard times are usually the best moment to take the opportunity to change
Let me give you more details about the changes, or rather the tidal wave of change, that Asset Managers can leverage to outperform the market in the coming years.
Picking the right security with Big Data
Today, finding the rising star has become very hard. With over $100 trillion of bonds on the market, 15+ currencies that are issuing bonds. Globalisation offers new opportunities every day to find the right issuer that would outperform the market but due to the huge amount of data involved, it is getting very hard for analyst. How can an individual compare the future performance of a Brazilian Public-Private Partnership bond issued in real (BRL) with a Midsize corporate bond from China issued in dollar (USD). There is too much data for the time available.
To me, the first challenge that technology could solve is to automate the data crunching and processing to make it easier and quicker for investment managers to make their decision.
This type of solution is usually developed internally by Hedge Funds (and sometimes Asset Managers) but it is not yet spread all over the market and so most Asset Managers are struggling to pick the bond that would meet their requirements. Indeed, it is very challenging to find Automation or Machine Learning specialists and only few Asset Managers can afford this.
Finding bonds through online marketplaces
The second challenge is about bond liquidity and sourcing. The bond market is complex (9 times more individual securities than the US stock market) and static (451 times less turnover than the US stock market).
For many years, the Bloomberg chat has been the most used trading system for bonds…
In the past few years, certain solutions have appeared on the market to create a central marketplace. This type of market place will transform the whole Bond market : more liquidity and more accurate pricing.
Getting the right value (and price) with AI
The third challenge goes together with the second. Buying the bonds before they get too expensive or selling it at a good price and avoiding fire sales…
Unlike stocks, bonds do not have a set value and in period of intense trading, the price volatility rockets and you’d better be on the right momentum when it happens.
To avoid these fire sales, AI could be the solution by predicting credit quality trends. By predicting this deterioration ahead of the market, this tool would help investors to sell bonds gradually and in advance, avoiding the drop in price.
Lowering the risks with Machine Learning
Last challenge, but not least: managing the risk of downgrade and default.
Statistics have been used by actuaires for years now in an attempt to lower the risk of default in the retail market of insurance, but a only few solutions are available at the Asset Management level.
Of course, there are solutions, that help risk managers to monitor exposure and risk. Yet, very few tools use Machine Learning to predict the trends of the market. And that’s where Artificial Intelligence brings most of its value, to predict and anticipate.
I am convinced, technology (and AI in particular) will disrupt the Asset Management industry. The question is not how but when. I believe the actual economic trends could accelerate this tidal wave of change. The next challenge for Asset Managers is to join this technological wave to outperform the market and avoid drowning in its wake.