A lot of time collectively is spent monitoring and analysing data on investment fund flows. Is this time well spent?
The assumptions typically made are that fund flows data indicate which assets are ‘popular’ and in the absence of significant change, continued flows might suggest underlying momentum for an asset class. Alternatively, very large inflows or outflows could indicate investors have become extremely positively or negatively disposed to an asset class, perhaps offering a contrarian investment opportunity.
An obvious initial challenge is whether data on backward-looking flows helps predict future flows. But there is a more fundamental difficulty with these arguments, as Jenny has discussed: at the market-clearing price (which in liquid public financial markets usually applies) supply and demand are equal. For every buyer, there is a seller (or net issuer). There is no net flow.
The fund flow data which is typically presented also only captures a sample from a segment of the potential investor base – mostly retail and institutional investment funds and ETFs. But what about everyone else? Think of central banks, sovereign wealth funds, hedge funds, individuals and corporates themselves. Once you include all investors, aggregate investor flows will sum to equal the net supply of new shares or bonds. The funds flow data as conventionally reported is just one side of the coin.
It is no surprise therefore that flows data provide limited explanation for concurrent market moves, let alone what will happen in the future. The chart below shows monthly US equity fund flows (including ETFs) over the last couple of years alongside the S&P 500. Recent large inflows have gone alongside price increases, but they didn’t in early 2015, nor did outflows in 2016 stop the market appreciation after the first quarter. A longer history reveals a similarly unstable relationship.
For signs of an abundance of capital flowing into a particular market or sector, suggesting risks of low future returns on capital, it would be better to analyse aggregate equity or debt issuance in that market.
To argue that there is information in funds flow data that is useful for predicting future investment returns, therefore requires a view that the characteristics and behaviour of the funds captured by the samples are in some way different to other segments of the market. Are they the smart or dumb money; weak or strong hands? Do the liquidity characteristics of the asset class sit at odds with the preferences of a particular subset of investor? Are the fund flows sampled particularly sensitive to benchmark risk or short-term market movements, perhaps susceptible to market-chasing behaviour?
Even if you are able to construct a theory of investor segmentation, some assumption is required around the relationship between flows and asset prices. Any such relationship is evidently loose. This we know because large price moves do not require any flows. It is easy to think of scenarios: market prices ‘gap’ in response to a shock event; or prices move in anticipation of a large flow (as described by Benoit Coure in a recent discussion of the impact of ECB asset purchases).
Or markets move because market participants just agree on a different price (i.e. beliefs change). The behaviour of US equity fund flows is instructive. Domestic US equity mutual funds have been in consistent outflow over most of the last 10 years, during which time the S&P 500 has delivered a total return in excess of 100% (June 2007-2017). As we know, US companies have been buying back shares. But this illustrates the point.
Asset prices are determined by beliefs, not by transactions. And the change in an asset’s price will tell you whether the ex-ante price was too high or low in light of new information and updated beliefs.
The conclusion we therefore draw is to spend most time analysing changes in price and valuation, and trying to understand prevailing consensus beliefs – about the nature of the world economy and attitudes to risk and return. Thinking about under what circumstances and in which direction beliefs are likely to be challenged and investor behaviour change, we believe is a more fruitful, albeit complex, exercise.