In biology, ‘dead zones’ refer to low-oxygen areas in bodies of water where microorganisms have used up almost all the oxygen to degrade large amounts of biomass. The accelerated growth of biomass can arise from an enrichment of waters with excessive nutrients such as phosphor. Our excursus in marine biology serves to illustrate recent developments in financial markets. With phosphor as a metaphor for quantitative easing (QE), the biomass growth is analogous to the asset price inflation of recent years and the rise of the passive ETF business. The oxygen, in our analogy, represents fund managers and traders who are willing to take active bets against a prevailing trend. And the key question is therefore: is there enough ‘oxygen’ now left in financial markets to weather an unwinding of QE and the risk of forced sales as ETFs are redeemed?
We already touched upon the consequences of unwinding QE in last month’s edition of ‘Thinking Ahead’ when we described the concept of asset price disinflation. But how potentially dangerous is the rise of ETFs in this environment? According to ETFGI, the total assets in ETFs and ETPs listed globally reached more than USD 5 trillion.1 Based on data from EPFR, assets under management (AUM) rose by about USD 4 trillion over the past ten years, of which USD 2.6 trillion are inflows, and equity funds have by far the largest share of this. Active mandates are still significantly larger than passive ones though, with AUM amounting to circa USD 20 trillion based on EPFR data. Across all markets, active mandates still dominate ETFs – based on the above data set, roughly 80% of all assets are still managed in active mandates. But this share has declined significantly from 90% ten years ago.
The importance of ETFs for financial markets has hence risen, but one might argue that the overall share is still fairly low and there is no cause for concern, in particular as our data set on equities covers only 50% of the global market capitalisation – the rest is likely part of e.g. separately managed accounts and individual stock holdings. However, since equity markets are vulnerable to the marginal seller, we believe that our data set still provides added value.
Generally, emerging markets (EM) equities seem most exposed to ETF redemptions with approximately 30% of the fund assets held in ETFs. Latin America has the largest share with more than 50% of all funds’ assets tied to ETFs (Chart 1). While active management is still dominant among funds investing in the Latin American region (ETF share is only at 15%), country exposure is usually gained via ETFs – in Mexico and Brazil they account for more than 80% of fund investments. This also holds true for emerging markets in Asia, in particular for South Korea (Chart 2).
ETF share in % of total net assets
Share of ETFs in country funds (% of total assets)
In developed markets, the share of ETFs is smaller. Asia-Pacific is notable in this regard with a share of almost 60%. While Asian investors generally seem to prefer ETFs over active fund management, the monetary policy of the Bank of Japan is another reason for the rise in Japan country ETFs – it constantly buys equities and REITs ETFs, which has propelled their share to almost 70%. Generally, however, we observe a similar trend towards higher ETF exposure in pure country funds whereas regional funds are still dominated by active fund managers.
This means that from the sheer size of the ETF market to the active fund market EM and Asia seem the most vulnerable to ETF redemptions.
A second important point for equity markets are the dynamics in flows between passive ETFs’ and active mandates. Over the past two years, the largest share of money went into ETFs, with the US and Western Europe experiencing a textbook-like switch from active to passive fund management. In the end, however, all of the regions are prone to the risk of a sudden withdrawal of ETF money which cannot be compensated for by the cash available in active asset management. Want an example? Let’s assume that traditional equity funds hold approximately 1.5% of their assets in cash. From January to November 2018, US equity ETFs witnessed inflows of about USD 130 billion (Chart 3). This is 1.7% of the assets of actively managed US equity funds. If market participants withdrew the money from ETFs to invest it outside the (US) equity space, the cash of active managers would not suffice to cover the redemptions. And this does not take into account the fact that active fund managers might not be willing to add to equities because of outflows similar to the USD 190 billion or 2.5% of AUM which were seen over the same time horizon.
It should be quite clear that an ETF sell-off would be detrimental to equity markets.
There is one more thing in fund flow data to which equity investors need to pay close attention: the rise of multi-asset funds. They are not as big as the equity fund industry – their AUM amount to ‘only’ USD 2.6 trillion. However, just like the ETF industry they too had benefited from significant inflows over the past ten years (with a short break between 2016 and 2017) given their implicit promise of positive and stable returns due to diversification and their ability to switch between bonds and equities as the environment demands. For the past six months, this asset class has seen remarkable outflows which even accelerated in the fourth quarter of 2018. While a rebalancing of these funds might have supported the drop in equities at the beginning of 2018, the wave of outflows observed in the second half of 2018 is likely to have brought equity markets down a leg lower. Clearly, further outflows would provide a giant headwind to equity market performance. And there is still fire power: of the almost USD 250 billion, which was added to balanced mandates since 2013, some USD 200 billion is still available for sale – including a performance gain of 40% over the past five years that has started to decline in 2018 by 2.6% (Chart 4).
ETFs have also gained in importance in fixed income markets. In particular, the share of ETFs in ‘plain vanilla’ mandates is high: more than 25% in the corporate bond space (for longer duration bonds even almost 60%) and in the government bond space (including inflation-linked and short-term government bond funds more than 40%). The share of ETFs in EM debt is comparatively low – about 10% (Chart 5). The question of how representative our data set is, becomes all the more important in the fixed income space. It covers about 5% of the global bond market, and 10% of the global EM debt market. However, the fixed income market – despite its size – is highly illiquid in the secondary market, in particular in the corporate bond space and the EM space. This means that despite the comparatively small share of ETFs compared to the total debt market, an unwinding of ETF positions is also likely to see larger setbacks in the fixed income market.
These days ETFs are a very important and integral part of financial markets. Their large share and the fact that there is no cash buffer in the case of redemptions could easily accelerate a downturn in risky assets. (EM) equity markets appear the most vulnerable given their comparatively high share of ETFs. Among bonds, the share of ETFs is much lower. However, the illiquidity of the (corporate) bond market makes even the small share of ETFs a risk one needs to be keenly aware of, while navigating the markets’ waters.