How to Avoid “Breaking the Buck”
I used to write about this topic after the financial crisis a lot. I stopped after the SEC made their last set of changes. But when I read this article today, I decided to tweet this:
Investment firms are wary as money-market regulation changes are studied after the run early in the pandemic. https://t.co/SS34ndPUeM My proposal makes more sense — NAV under 0.995, cancel enough shares to renormalize NAV to 1.0025, at which point runs strengthen the MMMF
— David Merkel (@AlephBlog) August 30, 2021
When I was an investment actuary, I developed and worked with a large number of insurance products that were managed to be pure savings vehicles. There was usually a short-term guarantee of crediting interest together with a full guarantee of principal. The difference between the short-term guarantee and actual earnings would get factored into the next short-term guarantee, making the short-term guarantee reflect the actual asset performance over time. Benefits would get paid at book value, while plan sponsor deciding to terminate would get something a little under the market value of the assets.
The names for those products were Initial Participation Guarantee, Guaranteed Investment Contracts, Stable Value Funds, and hybrids and synthetics of them. For fixed income investments that acted like savings accounts they worked well for pension plans and their beneficiaries in an era when interest rates started high, and kept falling. They were complex compared to running a money market fund because they ran with longer fixed income, making stability of principal much more tricky.
I even created a product that was relatively short duration that could trade like a mutual fund, making it eligible for a wide number of applications. (Sadly, I left shortly after it was created, and those who followed me didn’t run it right, and so it didn’t amount to much.)
So, when I write about money market funds [MMFs], I think I get the math going on there, and the problem is simpler because of the short duration, yet harder, because all assets are redeemable at par immediately. With the longer-dated insurance products, terminating customers would face a relatively harsh market value adjustment at termination. The challenge with MMFs is to make those demanding liquidity in a crisis pay the penalty, while those staying in do not get penalized, and possibly, get rewarded.
My proposal is slightly different than my prior proposal, but the math is the same. (Bond math is inexorable, even if bonds aren’t.) The first thing to do is define when a MMF is under stress, which is when its shadow NAV is under 0.995. If at the time the shadow NAV is calculated at the end of the day, the shadow NAV is under 0.995 then two things happen in this order:
- Those withdrawing funds lose additional units in the MMF in order to renormalize the shadow NAV to 1.0025. If any are doing a withdrawal of funds great enough that they won’t lose enough units, they suffer a market adjustment penalty of the same amount. The maximum amount of economic loss is 10% of the amount withdrawn.
- If the shadow NAV is still below 0.995 after step 1, units of those that remain are cancelled to the degree that the shadow NAV renormalizes at 1.0025.
Why penalize those who exit in a crisis? They are causing the crisis. Liquidity is not a free good, and certainly not in a crisis. This will make them think twice about liquidating, because they will absorb a disproportionate amount of the loss.
Why renormalize to 1.0025? Why not to 1.0000? This helps prevent further runs on the fund. If parties exit when the shadow NAV is over 1, it will boost the shadow NAV further. Even if the continuing holders lose units when the shadow NAV is renormalized, they don’t lose any value unless there are genuine defaults within the assets held in the MMF. If there are no defaults, the unit losses will be made up by a higher yield which will pay back the unit losses over the weighted average life of the assets involved.
A weak analogy to this is what Vanguard does on some funds that are less liquid. They charge a small termination penalty, but it gets paid to the fund, not Vanguard. Long -term holders get a small benefit from this.
The benefits of this proposal are:
- Allows all MMFs to have a $1 share price all of the time
- Penalizes those that grab for liquidity when an MMF is under stress
- Discourages runs on MMFs
- Would allow regulators to allow Prime MMFs to run a more aggressive investment policy
This sounds too good to be true, right? Well, it isn’t. My next post will run some scenarios to show how this would work. Personally, I think if this were adopted, corporations using Prime MMFs would appreciate the stable share price of $1, the potentially higher yield, but against that they would think about how liquidity would not be so easily available in a crisis, and that they would have to report credit losses when units are renormalized.
There is no free lunch, but there are ways to discourage people from running for the exits in the midst of a crisis, to the harm of everyone else.
PS — when I started writing tonight, I was just going to trot out my old proposals, but when I went through my history of working with guaranteed insurance products, I was reminded that our best risk-controlled products spread the losses onto those who leave at inopportune times. When I realized that, my old idea got a lot sharper.
Article by David Merkel, The Aleph Blog