Institutional asset managers can aggravate market swings due to the pro-cyclicality of redemptions, internal leverage and cash positions. A new empirical analysis shows that cash hoarding, a rise in funds’ cash positions in times of redemptions, is the norm. Cash hoarding seems to be particularly pronounced in less liquid markets and is a rational response if fire sale haircuts are prone to escalate with growing flows, i.e. if liquidating late is disproportionately costly. Investment opportunities arise initially from timely positioning and subsequently from the detection of flow-driven price distortions.

Morris, Stephen, Ilhyock Shim and Hyun Song Shin (2017), “Redemption risk and cash hoarding by asset managers”, BIS Working Papers No 608.

The post ties in with this site’s lecture on price distortions, particularly price distortions that arise from liquidity risk, and this sites lecture on systemic risk related to institutional asset managers.
On theory and empirical evidence suggesting that leverage in asset management is pro-cyclical view previous post here.
On the mutually reinforcing dynamics of EM fund flows and EM bond prices and the proc-cyclicality of both redemptions and discretionary sales of fund managers view previous post here.
On the incentives for end investors to rush for the exit in distress view previous post here.

The below are excerpts from the paper. Emphasis and cursive text have been added.

What is ‘cash hoarding’?

“If asset managers use their cash holdings as a buffer to meet investor redemptions, they can deal with redemptions without resorting to the sale of the underlying assets. Such behavior would be consistent with a ‘pecking order’ choice of actions where asset managers draw on cash first, and only start to sell the underlying assets if the cash runs out…However, if asset managers increase their cash holdings in the face of investor redemptions, they will need to sell more of the underlying assets than is strictly necessary to meet investor redemptions. We label this type of liquidity management as ‘cash hoarding’. Cash hoarding may potentially reinforce the impact of investor redemptions by amplifying the sale of the underlying assets.”

Our approach to distinguishing investor-driven sales and discretionary sales is based on comparing changes in cash holdings with the inflows and outflows of investors’ money…Suppose that the fund starts with no cash holding at the beginning of the period, but ends the period with a positive holding of cash, in spite of the investor redemptions. Then the positive cash holding at the end of the period can be regarded as the additional, discretionary sales undertaken by the fund.”

When are asset managers hoarding cash?

“We hone our insights by using a global game model of redemptions…The fund manager faces competing objectives when deciding how much of the underlying assets to sell in order to secure cash. Other things being equal, having more cash on hand allows the fund manager to meet redemptions more easily, thereby defusing investors’ incentive to run. However, other things are not equal. If the cash has to be secured by selling risky assets at fire sale discounts, future returns to staying invested are reduced, making redemptions more attractive. The fund manager’s cash holding decision reflects the tradeoff between securing enough cash to meet redemptions comfortably, but not selling so much that eventual fund returns are reduced.”

“If the fund manager liquidates [early] he faces an [initial] fire-sale haircut [or liquidity discount]…If he liquidates…[late] he faces an additional fire-sale haircut…There is feedback between liquidation and redemption. Increased anticipated redemptions give rise to higher ex-ante liquidation. However, higher liquidation has an ambiguous effect on redemptions…the endogeneity of the investor redemption decision necessitates weighing the cash hoarding decision of the fund manager against the reduced incentive to run on the part of the investor.”

“[In the formal model] cash hoarding occurs when the fire-sale haircut that applies to late sales is more than twice the liquidity discount that applies to pre-emptive liquidation. Thus, it is the relative discounts [between late and early liquidation] that matter for cash hoarding, rather than the absolute levels of the discounts.”

What is the evidence for cash hoarding?

“We examine a large dataset of global bond mutual funds to ascertain whether the portfolio decision of the asset managers conforms to the pecking order model where cash holdings are used as a buffer to smooth shocks coming from redemptions, or whether the asset managers engage in cash hoarding so as to amplify the fire sale of assets that results from redemptions.

Our sample consists of bond mutual funds investing globally…in both developed market bonds and EME [emerging market economies] bonds. We retrieved data from the EPFR database [for] every month from January 2013 to June 2016 [for] 478 global DM [developed market] bond funds, 104 global/regional EME international government bond funds, 105 global/regional EME local currency government bond funds, and 37 global/regional EME corporate bond funds.”

“In our empirical investigation, we find that cash hoarding is the rule, rather than the exception. Discretionary sales of the underlying asset tend to reinforce investor redemption-driven sales…We find that…discretionary sales in the middle of investor redemptions is the most common of all cases for each group of funds.”

“We find that the incidence of cash hoarding is more severe for those mutual funds that hold more illiquid underlying assets

  • For global DM bond funds, there is roughly 3 dollars’ worth of discretionary sales for every 100 dollars of investor-driven sales…
  • For global EME international government bond funds…there are 7 dollars’ worth of discretionary sales for each 100 dollars of investor-driven sales…
  • For the EME corporate bond funds… 10 dollars of discretionary sales are associated with 100 dollars’ worth of investor flow-driven sales.
  • For global EME local currency government bond funds… there are 13 dollars of discretionary sales for every 100 dollars of investor flow-driven sales.

For all four groups of bond funds, we find that the previous month’s fund returns increase the current month’s investor flows.”

“Finally, we find evidence of asymmetry between discretionary purchases and discretionary sales. The positive relationship between investor-driven sales and discretionary sales is stronger than the corresponding relationship between investor-driven purchases and discretionary purchases.”

What are the consequences of cash hoarding?

“Just as the procyclical leverage decision of banks tends to amplify the credit cycle, the cash hoarding by bond fund managers may amplify fire sales associated with investor redemptions.”

“In addition, redemptions by one group of investors may exert negative spillovers on remaining investors through the shifts in the composition of remaining assets from liquid to illiquid ones, as well as the marked-to-market changes in the value of remaining assets…The less liquid the underlying assets are, the greater are the spillover effects of investor redemptions to remaining investors, thereby exacerbating the selling pressures in a run-like episode.”

“Our findings raise questions about the way that asset sales interact with the strategic incentives underlying investor redemptions. Although the net asset value of mutual funds adjusts to changes in underlying market values, there are time lags in the adjustment.”

“In practice, the strategic incentives between fund managers are also likely to play a role in determining the market outcome. Such interactions across funds may inject additional spillover effects in which when other asset managers sell and market prices come under pressure, an individual asset manager may be tempted to join the selling spree.”

SHARE
Previous articleFake alpha
Next articleMeasuring non-conventional monetary policy surprises
Ralph Sueppel is founder and director of SRSV Ltd, a research company dedicated to socially responsible macro trading strategies. He has worked in economics and finance for almost 25 years for investment banks, the European Central Bank and leading hedge funds. At present he is head of research and quantitative strategies at Macrosynergy Partners.