Fidelity Back-test for 1-day vs. 7-day yields

Okay, so I was asked what if we had 1-day yield history, would it make much difference?

So, I took that challenge to create a back-test. Here is what I did:

1.) I started with MM Optimizer v2 and set it to run for 365 days ending 2/17/23

2.) I downloaded from fidelity.com daily and 7-day SEC  yields for four similar fidelity funds to the 4 general funds from Vanguard:

VMRXX -> FDRXX

VFMXX -> SPAXX

VUSXX -> FDLXX

VMSXX -> FTEXX

* I deliberately chose the entry level versions of the Fidelity funds for consistency. I know there are other share classes with lower ER and higher returns, but I wanted consistency over highest rate. This is a good time to point out that this back-test is not really the one to look at regarding Fidelity vs. Vanguard. That can probably be done straight off expense ratios as otherwise these funds are very similar, except that Fido appears to be using Repos even more than Vanguard.

3.) I updated the USGO% in the back-test sheet to match the values reported by Fidelity for these funds.

4.) My tool normally pulls straight from the Vanguard site, so I manually replaced the back-test data with the Fidelity fund yields in the My Tranch sheet (this populates all the other tranches yield data).

5.) I ran it for a NJ tax-base as I didn't want to introduce the complexity of the state-specific funds. They normally don't win often, so I don't think it makes much difference.

6.) Voila, I have an Optimize Study for the Fidelity funds that is in all respects that same as the Vanguard study. I executed this once for 1-day yields and once for 7-day yields

I produced two reports:

1.) A simple comparison of the results of the 1-day test vs. 7-day test of Fidelity, with a difference calculation This should be sufficient to demonstrate if there anything to learn from the back-test. Note that Tranch A2 is a simple swap between the Treasury Fund and the Muni Fund. For the pure test of Muni/Treasury swapping this is the most important test case.

MM Optimizer v2 - Fidelity-7day - 1vs7 Compare.pdf

2.) I provide a sample of the 7-day back-test for Vanguard funds with all the same parameters set.

MM Optimizer v2 - Vanguard 7 day.pdf

Bottom line, I don't think the 1-day yields matter much. The resulting calculated back-test yields are bit higher, but that is expected since rates have consistently risen during the test period so rolling 7 day yields are bit behind all the time. This doesn't change the value proposition of the swapping though. Also, the magnitude of the benefit did go up slightly across the entire test, but relative value is still about the same from one tax rate to another. High tax investors get more benefit and low tax investors really see no benefit and twice as many swap events.

For very high tax rates and only swapping between Muni/Treasury, the relative improvement was the greatest. The one day yields allow a $13/swap improvement with only 10 swaps required over the year.

For lower tax rates, the value of swapping actually decreased.