Substituting Precision for Accuracy: What the Bond Markets Can Teach Us About the Law of Unintended Consequences
By Paul Daley, Head of Fixed Info Lab – BondWave LLC
Paul Daley, Head of Fixed Info Lab – BondWave LLC
Can you mandate measurement without metrics?
I recently read the Securities and Exchange Commission’s proposed Rule 22e-4 concerning Open-End Fund Liquidity Risk Management Programs (LRMP) and Swing Pricing. As the 415 pages of this document will attest, the Commission put a tremendous amount of thought and effort into their proposal.
The end result is a requirement for open-end funds (mutual funds and ETFs) to design a LRMP with certain minimum requirements. The minimum requirements include using a variety of metrics to assign every fund holding to a liquidity category based on the number of calendar days or business days (some liquidity buckets use one and some use the other creating overlaps) it would take to liquidate the position near the last valuation assigned to the position.
This seemingly intuitive concept gets complex fast. For example, depending on the size of the position, portions of it can be assigned to several different liquidity buckets. The fund may be able to liquidate half of the position in three days, but need another three days to liquidate the remainder. Also the same security may be considered liquid by one fund and illiquid by another fund based solely on the fund’s size. The security can move between liquidity buckets based on changing perceptions about future volatility, bid/offer spreads, or depth and quality of market makers.
“As important as liquidity management is for funds, liquidity promotion may be more important for the SEC”
Many of the objections to the complexity of the proposal are covered in detail by the comment letters the SEC received so I won’t go into any further detail here regarding them. Comment letters from BlackRock, Vanguard and the Investment Company Institute are detailed and well thought out and should be read by anyone interested in this topic.
Instead, I will focus on what I believe is one of the logical outcomes of this proposal if it passes unaltered. But it is not an outcome the Commission seems to have considered. Specifically, can the SEC impose a fund liquidity requirement that is granular down to the security level without also imposing a liquidity transparency requirement for fixed income markets?
Among the SEC recommended metrics for calculating the number of days to liquidate a fixed income position are: number of quotes, the number of dealers willing to purchase or sell a security, the number of other potential purchasers, volatility of trading prices, and bid-ask spread for the asset. The problem arises from the fact that for some securities in the current market structure complete, accurate measurements of these metrics are near impossible, especially at the individual security level as mandated. For example, while it is possible to calculate an implied bid-ask spread for a category of bonds by looking at a subset of just those that appear to trade on the bid and the offer in close time proximity, this is at best a rough approximation for a market rather than an accurate measurement for each security.
By failing to acknowledge the lack of accurate data for these metrics the SEC seems to be demanding precision rather than accuracy. Of course, this can be harmful to investors who only see the final precise answer, but do not understand the guesswork that went into producing it. The precision of the answer implies an accuracy that is just not there.
Precision vs. Accuracy
A great analogy comes from the world of science where accurate measurement is critical to uncovering truths. In science, the concept of significant digits is used to make sure significance is maintained throughout calculations. Spurious digits are considered to be those that are introduced by calculations carried out to greater precision than that of the original data. In effect, the SEC is proposing the maintenance of spurious digits, rather than their removal, by insisting on using metrics that do not have the transparency to support accuracy.
A Transparent Solution?
Of course, one solution to the problem is within the SEC’s purview to implement. They can require pre-trade transparency for all fixed income products. It is currently impossible to know with a degree of certainty what the best available bid or offer is for a bond. Even with many technology-based attempts to solve the problem the search costs remain too high. Meanwhile, pre-trade transparency is required for equities, and futures effectively achieve the result by requiring all trades to occur in a single location (physical or virtual).
There are other benefits to requiring pre-trade transparency, too. Doing so would increase liquidity. Why? It is human nature avoid an activity if you are uncertain of the rules governing that activity. In 2004, while conducting a study of the municipal market, Lawrence Harris and Michael Piwowar found that unlike in the equity markets where the unit cost of trading increased with trade size, small retail trades were substantially more expensive than large trades. They “attribute the difference primarily to the lack of transparency in the markets.” In short, retail investors receive the worst outcomes. And it is a natural consequence that they would no longer want to play the game, thereby decreasing liquidity in the market.
This approach would also broaden participation, creating a more efficient market that would lower prices and trading costs. A 2015 study by Lawrence Harris and Fred Keenan of the corporate bond market reveals that 42.9 percent of all transactions occurred at prices worse than the best bid or offer available at the time of the trade (“trade through”) and that 96.8 percent of the trade throughs occurred in two sided markets. Because these calculations were based on an aggregation of quotes from only five different electronic trading platforms they represent the lower bounds and would most likely be higher still if they had access to all quotes (pre-trade transparency). In the current market structure, even institutional traders cannot be confident in getting or paying the best available price for their bonds. In addition to promoting confidence which leads to greater usage, pre-trade transparency also promotes competition leading to tighter spreads and lower trading costs. Lower trading costs always promotes greater liquidity, and the effect is felt even more keenly in the current low interest rate environment where there is little yield to overcome trading costs.
While it is likely the SEC’s LRM proposal will go through a series of revisions based on the comments they have received from leading firms in the industry it is good to see a detailed expansion on the current thinking of the Commission. It serves to point out a fundamental issue in the current market structure. In the equity world, through Regulation National Market System (Reg NMS), the SEC treats liquidity as a public good by enforcing competition among venues to lower the cost of trading and ban trade throughs. And while we may debate their methods for achieving this goal and the unintended consequences, their theory of regulation is correct. However, in the fixed income world, due to the lack of pre-trade transparency, liquidity is not treated as a public good, trade throughs are allowed and high trading costs are generally accepted. As important as liquidity management is for funds, liquidity promotion may be more important for the SEC.