Are some traders gaming the system via settlement failures? | FT Alphaville

Are some traders gaming the system via settlement failures?

How worried should investors be about rising settlement fails?

In the opinion of the Kauffman Foundation, very.

A new study penned by among others Kauffman’s chief investment officer, Harold Bradley and vice president of research Robert E. Litan — a regulatory expert who has supervised financial investigations into NASDAQ and others — raises the alert on what they claim could be a canary in the coal mine in terms of systemic risk for the industry at large.

While the report is likely to prove controversial with some market practitioners — it certainly raises points which are worth investigating further.

Among them, settlement trade figures which expose a growing failure rate in two markets: mortgage backed securities and exchange traded funds (ETFs).

Both trends have already piqued our interest here on FT Alphaville.

The authors note that in spite of the Treasury Market Practices Group imposing a fining system in the Treasury market to discourage fails, and the SEC introducing a T+4 settlement deadline in equities, the problem refuses to go away:

It simply has moved to markets where fails are not punished. For example, the Treasury Department‘s primary dealers reported that in all but five weeks during 2010, sellers of Treasury, MBS, agency, and corporate fixed income securities failed to deliver each day an average of $130 billion of securities on the expected settlement date, a number so large that it exceeds the combined $89 billion regulatory capital of these institutions. During the week of November 24, 2010, daily fails of these securities exceeded primary dealer regulatory capital by more than two-and-a-half times.

In the same week, the primary dealers reported to the Federal Reserve that total MBS fails exceeded $1.3 trillion, an amount never before recorded in the more than fifteen years the Fed has collected data. This is the third time fails exceeded $1 trillion in less than six months during 2010, which continues a pattern of increasing MBS fails that has evolved since May 2009 when a penalty was put in place to stop persistent fails of Treasury securities.

It is impossible to know without more compulsory data reporting by the primary dealers and custodians exactly what accounts for these patterns, or whether regulators even are aware of the problem or its causes. The systemic risk to these dealers‘ liquidity is evident from the sheer size of the numbers and the potential impact on any firm‘s ability to continue operations if it were forced to honor commitments in these transactions during another crisis in either liquidity or counterparty confidence.

Meanwhile, with regards to ETFs — the problem is not so much related to the sheer size of the settlement fail issue but rather its ongoing nature.

Another major liquidity problem may be simmering given the rising frequency of fails in ETF securities. Currently, ETF fails account for approximately 60 percent of the nearly $2 billion of daily equity trading fails reported to the SEC, and on some days they account for 90 percent of all exchange traded fails.

Understanding the cause of ETF fails is even more complex given the number of intermediaries involved in the creation, marketing, sale, pricing, and safekeeping of ETF securities. Mandatory reporting for each of six legal entities involved in the daily maintenance of each ETF security would assure that investors understand the inherent risks in the wide range of ETF construction and trading strategies.

Turns out it is often the largest ETFs (SPY, XLF and XLI) that have the highest failure rates,  according to the authors — sometimes reaching over 240 days of continuous failure:

So why would ETF and MBS fails be so much greater than those in respective equity and Treasury market.

Much of it, the authors agree, is down the two markets being exempt from the same kinds of penalties that regulators have imposed elsewhere.

That regulatory oversight has consequently led to the following  fail statistics:

Of course, others might argue those figures are misleading since they are taken against market value rather than compared to settled statistics.

But, overall, it is clear from the below chart that the trend is rising across the securities space:

While the ETF fail figure is low with respect to the share of the total equity market, there is still reason to be concerned, according to the authors:

While ETF failures are magnitudes of order smaller than MBS failures, they have the possibility of being the first in a string of dominoes to fall in a crisis. ETFs are highly visible but their failures provide one of the best examples of ―things not being as they seem.‖ ETFs are marketed and sold as exchange traded equity securities that have all the diversification advantages of a mutual fund with unlimited supply and they can be traded (and shorted) throughout the day. Investors have been lulled into the belief that ETFs are just like equities through repeated assurances from brokers and the issuers.

For this reason, they conclude, the current failure levels present signifcant systemic risk to all investors in the event of another market shock.

One also has to wonder about the current incentives involved —  why are these failures happening in the first place?

The authors speculate thus:

If the failed trades result from intentional failures to settle, then the authorities are letting parties game the system at the expense of the beneficial securities owners such as state pension funds, mutual funds, and retail investors, who all suffer economically as a result of this behavior.

But, again, not all would necessarily agree with this point.

In the bond market — especially in a low interest environment — there is not always an economic interest in pursuing claims on fails, which are often the result of pure operational or technical issues. That said, in normal times fails can be seen as an accepted market practice. One side costs the other money, a sum which is later reclaimed on a ‘cost of money’ basis.

This is understandable because the size of transaction is so much greater in bonds, and a few basis points can make a very big difference.

In equities, though, ever declining trade sizes  generally mean it is  less economically viable to ensure settlement on every single transaction taken (or so we were told).

Furthermore, it doesn’t make sense that participants would be failing on purpose —  for example, to use the securities elsewhere.

First, a market participant begins to accrue the profit and loss (including from dividends) from the moment of execution, not settlement. Second, it would be the custodian who would be the beneficiary or victim of an unsettled trade, not necessarily the customer.

That said the issue of the ETF failures remains puzzling.

One market source told us:

It is entirely possible that there is a financial incentive to fail trades but we need to quantify this. Why are ETFs exempt from failure penalties at the moment. What would be the incentive for a market participant to fail?

After all, if I was an investor in an ETF and was expecting a delivery and the counteparty failed, I would have a claim on the creditors which could now be worthless. At that stage the only way I can sell is if I get a delivery.

Of course, if you give up on the prospect of that security ever being delivered from the administrators of a defaulted party you might choose to close out your trade via an alternative. Say via the futures market. The fail itself, however, will continue to live on, cascading from day to day and so on. A fact which makes our source wonder if some of these long-lasting ETF fails are a residual factor from the Lehman bankruptcy?

Either way, this does remain a puzzle. And it makes sense to solve it — especially if somebody somewhere is making money continuously at someone else’s expense.

For more, check out the full report here.

Update: 18:24 GMT – Index Universe’ Dave Nadig has since provided one logical explanation —  ETF market-makers are entitled to T+6 settlement, while the National Securities Clearing Corporation “fails” report counts anything over T+3 as a fail, even if it is technically not.

Related links:
What’s the ETF settlement fail issue?
– FT Alphaville
That cornered – failing – MBS market
– FT Alphaville
Repos turning Japanese
– FT Alphaville