Term SOFR Rates: Let’s NOT allow a bona fide new benchmark to be polluted before it gains precedence.

As always, my thoughts and opinions are solely my own and do not reflect views of any employer, employee, or any person or entity. Nothing herein is trading, accounting, tax or any form of advice.

Why pretend like term rates are hard to derive? —UNLESS, some entities are trying to find a way to try and sell Treasury risk at a much more profitable pricing scheme than selling Treasury securities or Treasury repo? What if we renamed tap water, ‘Super-Fluid’, and repackaged it, using a new liquid measure that was neither milliliters nor ounces? And, got some government to say, “Super-Fluid is the key to life!”. Folks, it’s just water. Ok, I’m officially asking, “Is someone looking to make SOFR term rates the ‘Super-Fluid’ of finance”?

No ‘buts’ on SOFR: I love SOFR and fully believe in the transparency, the rate efficacy, and utility for RISK-FREE transactions. I do not like, at all, the discussions of ‘cloudiness of term rate’.

“SOFR is very volatile!” No kidding. Were there no repo or arb traders on the ARRC? Treasury repo is, and always has been, very volatile. It’s about to get worse with wild swings in T-Bill issuance due to (at some point) resolution of the government shutdown. Then, there will be more swings as the assured-to-be-bombastic-and-drawn-out budget deadline approaches, then the push-forward, blah blah blah. Along with ebbs and flows in the Treasury repo market, T-Bill issuance (or lack thereof) also causes significant volatility. Always has. Always will. Nonetheless, there are plenty of market makers for term repo. AND, there are plenty of market makers for 1mo and 3mo term SOFR futures (for a couple YEARS+ already). How can that be with the term SOFR rates sooo difficult to render?

Treasury TERM repo is, and always has been, calculable and trade-able because of ‘NO-ARBITRAGE’ pricing via cash flow discounting and parity of identical credit and term Treasury instruments. Almost as easy as it is to find a term Treasury rate, it is to find a Treasury term repo rate. And, since SOFR is an overnight Treasury Repo hybrid, term SOFR is term Treasury repo. And, term Treasury repo is a function of a term Treasury security’s rate. (Spoiler alert: if you can find a term Treasury rate, you can find a term SOFR rate). The only problem with finding a term SOFR rate, is figuring out how any swap dealer can make money when the competing product (a Treasury future or cash) offers virtually no profit opportunity.

Somebody wants a 10-year SOFR rate (i.e. ‘a 10-year Treasury term repo rate’)? No problem. For those that don’t run arbitrage or discount cash flows for a living, here’s the example of “No-arbitrage” pricing in terms of 10-year SOFR/term repo vs 10-year Treasury Security:

Example: Buy $1mm 10yr Treasury. Pledge $1mm 10yr Treasury as collateral. Borrow $1mm for a term of 10 years (term repo aka term SOFR). This trade, buy or sell, MUST RESULT IN NO PROFIT FOR EITHER SIDE OF THE TRANSACTION (absent cost, haircut, etc).

The sum of the present value of the cash flows of the 10yr Treasury Security (‘the collateral’) must be equal to the sum of the present value of the cash flows of the loan (10yr Term Repo).

This IS No-Arbitrage pricing. If the borrower and the lender have the same risk, neither side can make a profit—or every borrower would become a lender (or vice versa) until there was no more ‘FREE MONEY’.

TERM RATES aren’t always right (that’s why they change constantly). Would you rather own a 6-month T-Bill and roll it for a decade, or own a 10-year semi-annual pay Treasury? The answer is, “Totally indifferent—the answer won’t be known ‘which is better’ until after time has passed.” Of course, just as T-Bill rates won’t stand still for a decade, neither will the price of the 10-year as it approaches maturity. BUT, ‘No-Arbitrage’, means ‘instantaneously, economically indifferent’. If there was an economic advantage to rolling 6mo T-Bills versus owning a 10yr semi-annual pay Treasury, everybody would own the advantaged choice and avoid (or sell short) the disadvantage choice. No arbitrage.

Of course, a SOFR RATE can be different than an equal maturity Treasury RATE; due to differences in compounding frequency. An overnight rate for X-years, differs from a 1-month rate for X-years, differs from a 3-month rate for X-years, differs from a 6-month rate for X-years, and so forth… BUT, due to ‘No-Arbitrage’, the SUM OF THE PV OF CASH FLOWS MUST BE IDENTICAL—whatever the interval of compounding.

Arguably, we don’t need TERM swaps on SOFR—just trade everything on spread to equal maturity Treasury (the equivalent of trading solely the fixed leg of a swap). Need fixed SOFR funding? Buy a bond, pay SOFR+X overnight to finance it, short an equal maturity Treasury, receive Treas repo overnight, pay Treas fixed rate coupon. No swap needed. The overnight Treas repo receive (from the short Treas or T-future) washes with SOFR. Cost of funds is Treas Coupon + X. It's even easier using Treasury futures without risk of a squeeze on cash T short. The point is: no swap needed. With SOFR as a funding rate, a short Treasury Future/cash + repo IS the swap. No need to incur additional cost. BECAUSE, anyone can be a borrower or lender, overnight or for term, in the Treasury repo (aka SOFR) market. It's collateral, not credit, that determines participant eligibility.

Unsecured, interbank-only, credit-based, LIBOR needed a swap market because there was no ability to borrow and lend at LIBOR for the general public-- it was a polled and published interbank, credit-secured rate. Yes, anybody can mark off LIBOR-- but everybody cannot actually borrow/lend at the rate. In the case of the Treasury-secured repo market-- borrow or lend, all you want, massively competitive, every business day. No swap needed--just Treasury collateral or cash. Candidly, I'm shocked the dealer-bank community let this go through. Unless the reps on the ARRC weren't people with vision into this subtlety.

I get it that no one wants to admit that TERM SOFR is a re-invention of the Treasury market… eh, well, uh… No one who makes money trading swaps wants to believe LIBOR is being replaced by something that does not require swaps trading—or worse, a swap that trades a bid/ask like a Treasury and must be collateralized as a cleared derivative with internal costs. For those with Bloomberg, you may have noticed the YAS function. It’s not a new invention in the last 20+ years. YAS shows yield spreads on many non-Treasury bonds relative to the ‘closest to equal maturity active Treasury security’. Spread to Treasury isn’t a new idea. It simply lost out to LIBOR because of credit sensitivity. LIBOR had credit sensitivity. Treasuries did not. Back to the future we go.

Since ‘credit’ has been wiped from the new benchmark risk profile, all of the volatility between a non-Treasury and equal-term SOFR will be manifest in increased volatility in the fixed-to-fixed (Term SOFR OAS) spread volatility (until a Credit-SOFR product/process is created: https://www.dhirubhai.net/pulse/re-benchmarking-sofr-ameribor-fill-two-silos-proposal-john-coleman/ ). The term SOFR quandary could be handled in simple spread-to-Treasury pricing and trading. (Don’t forget to re-benchmark all non-Treasury OAS models while changes are being made to systems. Ugh.)

Here’s one obvious answer: Have the exchanges publish (as is done for cleared LIBOR/OIS swaps) a schedule of daily discount factors for the entire Treasury curve. Interpolate a zero curve. Calculate discount factors REAL TIME, distribute (yeah, ‘sell’) the data. No matter your start/end dates, coupon or payment periodicity, everything can trade off the interpolated equivalent maturity Treasury. Build amortizers, accreters, roller-coasters—whatever is needed, from the real-time Treasury discount factor curve. Everything can settle to that curve. And, there’s an immensely liquid and developed, ALL-TO-ALL, Treasury futures (and a Treasury cash) market, with OPTIONS ready to go. Heck, have the Fed publish the discount factor curve’s daily settle on the H15 report every day. Trade 1mo and 3mo SOFR futures out to 2-years. Then, let the EXISTING Treasury market and No-Arbitrage pricing take over. Or, maybe the best way to pollute ‘all that is right with SOFR’, is to push the “Super Fluid” pricing obfuscation on term rates.