The Big Idea
The evolution of liquidity and its fair value
Steven Abrahams | August 13, 2021
This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.
The repo market prices liquidity every day, and the picture it currently paints is one of still healthy compensation for less liquid assets. The market in private MBS has become more liquid over the last year while private ABS has become less liquid. But selling liquidity is still at attractive source of return in a market with limited sources of yield.
Using the repo market to price liquidity
Investors can know a lot of things about securities sitting in a portfolio, but liquidity is one of the most elusive. There is trade volume, bid-ask spread, the price impact of big trade flows. Each provides part of the liquidity picture. The New York Fed provides daily information about trading volume across a wide set of assets. Quoted bid-ask spreads show up for some assets but are estimated or largely anecdotal for most. And the price impact of large trade flows is perhaps the most difficult to measure of all. But there is another indicator of liquidity. In fact, it is an entire market.
The repo market traffics in liquidity every day and implicitly measures it by the haircut or margin required for different assets. Repo desks take in assets as collateral against loans, and those desks have to understand their ability to seize and sell assets to recoup the loan, especially in a distressed market. In the deepest and most liquid markets, the time to sell, the bid-ask spread and the price impact of even large flows is small, In thinner markets, the time to sell is longer, the bid-ask spread wider and the price impact larger. Repo haircuts reflect differences across assets in all these factors.
Opinions on asset liquidity will differ across repo desks and so will haircuts. For markets with standardized assets and transparent pricing, most repo desks will have a similar view and dispersion in haircuts will be small. In markets with differentiated assets and opaque pricing, views will differ, and dispersion will be wide. Both the average or median repo haircut and the dispersion in haircuts carry information about asset liquidity.
The latest picture from the NY Fed
The New York Fed provides a rich picture of repo levels and dispersion across assets, and that gives a monthly snapshot of implied liquidity. The latest picture shows the usual assets falling into three broad categories (Exhibit 1):
- Higher liquidity. This includes US Treasury bills, notes and bonds, US Treasury STRIPS, agency debt and agency MBS. All of these have median haircuts of 2% with dispersion across repo providers of less than 2%, meaning two-thirds of providers offer haircuts between 0% and 4%.
- Intermediate liquidity. This includes money market debt, IG corporate debt, IG private-label CMOs, international securities and agency CMOs. All of these have a median haircut of 5% except for agency CMOs, which have a median haircut of 4%. But the dispersion in haircut varies significantly. Money market haircuts have a dispersion of 1.34% while agency CMO haircuts have a dispersion of 4.99%.
- Lower liquidity. This includes equity, non-IG corporate debt, muni debt, non-IG private-label CMOs, and IG and non-IG ABS. All of these have 8% median haircuts, although IG ABS has a 7%. And again, the dispersion of haircuts across assets varies significantly.
Exhibit 1: Median and standard deviation of asset haircuts for July 2021
Some changes from a year ago
The liquidity picture has changed only in some specific instances over the last year, with liquidity broadly improving in 2021 with a few exceptions (Exhibit 2). The private-label CMO market has become notably more liquid since 2020, for instance, with the median haircut on IG classes coming down and dispersion dropping. The median haircut on IG private-label CMOs has dropped from 7% to 5%, and dispersion has dropped from 4.4% to 3.2%. In non-IG private-label CMOs, the median haircut has stayed at 8.0% but dispersion has dropped from 5.6% to 4.4%. In contrast, IG ABS has become less liquid with dispersion rising from 4.5% to 5.7%, and muni debt has become less liquid with both the median and dispersion rising.
Exhibit 2: Median and standard deviation of asset haircuts for July 2020
Summarizing with a liquidity score
The relative order of liquidity and the change over time becomes a little clearer by calculating a liquidity score for each asset—simply the sum of the median haircut and the dispersion. A low score indicates more liquidity. Comparing liquidity scores across assets for the most recent snapshot order assets from Treasury bills, notes and bonds on the most liquid end to non-IG ABS on the least liquid (Exhibit 3). And comparing liquidity score in July 2021 to July 2020 shows most assets have become more liquid—their scores have fallen—while ABS has become less liquid.
Exhibit 3: Liquidity scores show most, least liquid assets and change over time
The fair value of liquidity
Although the fair value of liquidity is more art than science, one angle starts with the spread on ‘AAA’ corporate debt. That spread is more for liquidity than for credit. Moody’s studies of corporate credit show that since 1970, the average ‘AAA’ corporate bond after 10 years has a 0.37% chance of being in default with an ultimate recovery of 47.7% for an expected loss of 0.18%. That should be the expected compensation for credit loss. Readily available spreads on ‘AAA’ debt go back to 1996, where the ICE BoA AAA Corporate Index OAS shows a median of 0.68%. It seems reasonable to assume that 26% of the typical ‘AAA’ spread is for credit risk and 74% for liquidity risk. With today’s ‘AAA’ spread at 52 bp, the fair value of the liquidity difference between ‘AAA’ corporate debt and Treasury debt is 38.5 bp.
Based on liquidity score, the 38.5 bp of liquidity premium for ‘AAA’ should scale up or down for each asset class, assuming the fair value for Treasury liquidity is zero. Depending on asset class, a substantial part of current spreads to the Treasury curve are compensation for the asset’s lower liquidity (Exhibit 4). In a market with generally tight spreads, compensation for liquidity in may credit assets looks attractive.
Exhibit 4: The fair value for liquidity ranges significantly across assets
With Fed tapering in the mix for the balance of the year, the flow of liquidity into the market looks set to decelerate. Liquidity and compensation for liquidity should plateau. But with liquidity risk still a significant multiple of compensation for other risk in credit assets, it still looks like an attractive source of return in a market for limited sources of yield.
* * *
The view in rates
Fed RRP balances closed Friday at a near-record $1.05 trillion, up roughly $100 billion from last week. At 5 bp, the rate on the RRP is still attractive relative to other forms of repo.
Settings on 3-month LIBOR have remained steady over the last week at around 12.5 bp, still near the lowest setting ever. LIBOR has started pricing in the transition to SOFR. Most interdealer swap trades will use SOFR by October at the latest, with regulators urging no new LIBOR exposures after the end of this year. LIBOR officially sunsets on June 30, 2023.
The 10-year note has finished the most recent session at 1.28%, down 1 bp on the week. The yield ran slightly higher until a weak reading on University of Michigan consumer sentiment brought it down. Breakeven 10-year inflation is at 238 bp, up on the week by 2 bp. The 10-year real rate finished the week at negative 110 bp, down 4 bp on the week. Real rates have improved since touching negative 120 earlier this month but remained bogged down by concerns over an excess supply of liquidity and insufficient growth.
The Treasury yield curve has finished its most recent session with 2s10s at 106 bp, 3 bp flatter than a week ago. The 5s30s curve has finished at 116 bp, flatter by 1 bp.
The view in spreads
Credit spreads have softened a few basis points lately. Benchmark investment grade cash spreads are only wider by 3 bp from the end of May and should continue to outperform. Demand from mutual funds, international portfolios and insurers looks healthy. Low rates should continue supporting corporate balance sheet strength. Ratios of EBITDA to interest expense are in the middle of the range despite high ratios of debt to EBITDA. Investor demand for yield should keep spreads relatively tight. A strong economy should help credit spreads, but relative value flows at money managers could still soften credit spreads if MBS gets wide enough.
MBS is finishing Friday at its widest levels since March. The prospect of Fed tapering and heavy net supply should keep weighing on MBS spreads until the Fed shows its hand and the market can price the impact. The market has already priced additional risk of soft demand and steady supply, with the nominal spread of par 30-year MBS to the 7.5-year Treasury at 75 bp, wider from the end of May by 13 bp. Spreads look vulnerable to going still wider as the Fed likely leans into tapering or tapering-and-hiking faster than its 2013-to-2015 cycle.
The view in credit
Fundamental credit looks strong and generally continues to improve, helped by Covid reopening and low rates. Consumers finished the first quarter of 2021 with net worth up $5 trillion. The second quarter should add to consumer net worth with equities and real estate higher and rates even lower. Consumers have not added much debt. Corporate balance sheets have taken on more leverage, although mitigated by strong cash balances and low interest costs. EBITDA-to-interest-expense is at healthy levels. Strong economic growth in 2021 and 2022 should lift most EBITDA and continue easing credit concerns. Eventually, rising interest expense in 2023 should compete with EBITDA growth. Fundamental credit should hinge on whether the Fed can orchestrate a soft landing as it starts to tighten financial conditions.