During the fourth quarter of 2018, US high yield bonds suffered a negative total return of 4.5%. Also during those three months, investors pulled a net USD 13.8 billion from US high yield exchange-traded funds (ETFs) and mutual funds.
Evolution has trained our brains to be excellent pattern-recognition machines, so it's natural to feel the urge to connect these two data points, which would tell the story that investors pulled assets out of high yield bonds and therefore caused the asset class to lose value.
But there is no causality linking these observations. This is one example of a logical fallacy, "cum hoc ergo propter hoc" ("with this, therefore because of this"). In fact, fund flows tell us very little, for two main reasons:
- First, as we discuss in The myth of "more sellers than buyers," all market trades are an exchange of ownership between investors. Fund flow data does not tell us that cash has entered or exited an asset class, only that more of that asset class is being owned in a given structure (for example, ETFs).
- Second, ETF and mutual fund assets represent only a slice of the overall market, so fund flows into and out of this slice only tells us about this small part of the overall picture.
Let's take US high yield funds as an example. Only about 22% of the asset class is owned through ETFs and mutual funds; the other 78% is owned through other third-party managed investments (separately managed accounts, hedge funds, etc.) or individual bond holdings (see Fig. 1).
Fig. 2 shows us that there has been a sizable outflow from high yield mutual funds over the past decade—a trend that has mostly been offset by flows into ETFs. These structural trends illustrate investors' evolving preference for how to own high yield bonds, but no insight into investors' demand for the asset class overall. Even if you knew the exact amount of high yield inflows occurring next week, it would tell you precisely nothing about what the asset class's return might be. Regardless of the net outflow from mutual fund and ETF assets over the past decade, the US high yield bond index has delivered a total return of 200.8% (11.5% p.a.).
By the same token, looking at past flow data doesn't help us better understand the mindset of investors. All investors have access to multiple ownership structures, so it's impossible to separate "smart money" from "dumb money"; instead, flows can only give us a glimpse of the shifting landscape of how investors are choosing to own a given asset class. From an individual perspective, we may have a preference for whether to own high yield bonds in ETFs or mutual funds or separately managed accounts—but market returns aren't significantly affected by that breakdown.
Fig. 1: Only a sliver of the US high yield market is owned through ETFs and mutual funds
Fig. 2: Within US high yield, ETF and mutual fund flows have diverged
Author: Justin Waring, Investment Strategist Americas, UBS Financial Services Inc. (UBS FS)
David Perlman, ETF Strategist Americas, UBS Financial Services Inc. (UBS FS)