In the wake of the market frenzy around GameStop, some stakeholders such as the Depository Trust Clearing Corporation, are examining whether counterparty risk in stock market transactions could be reduced through a shortening of the settlement period from the current two-day (T+2) market standard. A move from T+2 would also affect the securities lending market, which provides the plumbing needed for short selling. In this note, we assess the potential impact by evaluating what happened in the securities lending market when the settlement period was reduced from T+3 to T+2 in September 2017.
We find evidence that brokers, custodians and securities providers (e.g. mutual funds), who collectively provide the servicing and supply in the securities lending market, were able to continue providing the liquidity needed to make the securities lending market function without disruption after the 2017 change from T+3 to T+2. However, a further shortening from T+2 to T+1 could cause more frictions than the 2017 change, depending on servicers and suppliers continuing to have adequate time within 24 hours to execute the multiple transactions across various time zones that would need to occur for seamless delivery and settlement.
To proxy the liquid functioning of the securities lending market, we examine three metrics before and after the 2017 shift from T+3 to T+2:
Failures to deliver occur when a party to a transaction does not settle and exchange securities for cash in a timely fashion. Fails can occur by force or by choice: that is, a party may not be able to come up with the securities in the specified timeframe (very rare) or the party may strategically decide that the penalty for failing is lower than the cost of delivering (also very rare). If the September 2017 reduction in settlement from T+3 to T+2 had been disruptive to the securities lending market, we might see a rise in fails. The rationale is that brokers might have needed a full three days to “locate” a stock for purposes of fulfilling a short sale, and reducing the locate period by 24 hours could prompt an increase in fails. In fact, by 2017 the locate process has become highly automated and efficient, and consistent with that, the move from T+3 to T+2 appears to have exerted no discernible effect on parties’ ability to deliver shorted stocks in a timely fashion, as the Chart 1 below illustrates.
No Data Found
Another potential indicator of friction in the securities lending market would be evidenced by a reduction in the willingness of market participants to short stocks. For example, if a reduced settlement period translated into less supply of stocks available to borrow for shorting purposes, we would expect to see aggregate short interest decline in response. As Chart 2 illustrates, aggregate net short interest declined slightly in 2017, but this was more likely attributable to the sanguine outlook for the stock market in the early period of the Trump Administration—making shorting less attractive. Moreover, the level of aggregate net shorts after 2017 remained within ranges experienced since 2010.
No Data Found
We find that, when the threshold is stocks with net short interest of at least 30 percent, there does appear to be a slight decline in short interest beginning in 2017, but again the level of short interest on these highly shorted stocks is well within historical ranges and is likely attributable to other factors. When the threshold is set at 50 percent to capture only those stocks that are very heavily shorted, aggregate short interest is stable to slightly higher after 2017. While the relationship is not clear cut, this is consistent with some possible concentration of shorts in a narrower set of companies—meriting further examination.
No Data Found
In sum, the shortening of the settlement period from T+3 to T+2 in 2017 does not appear to have had much impact on the stock lending markets used by short sellers. However, in surveying prime brokers on the potential impact of a further reduction to T+1, some market participants indicate that the 24-hour turnaround time could lead to a slight decline in both the supply of securities for shorting and in the willingness of brokers to allocate borrowed shares as liberally as they are accustomed to doing under T+2. In addition, there is an important question on whether all markets, including Treasuries, would need to move to a T+1 cycle at the same time in order to avoid disruption in markets.