Case Study: Identifying Value in Short-Term Floating-Rate Funding

Hi, I’m Tyler Buckridge with the Federal Home Loan Bank of Boston.

I want to anchor this case study around one idea.

If a short-term borrowing keeps rolling, it is already floating economically.

A one-month fixed advance may lock the rate for 30 days, but every maturity hands that balance back to the market at a new rate.

So what we’re really trying to understand is not simply fixed versus floating.

It is how that funding reprices, what certainty is worth, and whether the structure actually matches the life of the funding need.

That is the lens we’ll use throughout the case study.

We will move from SOFR mechanics to short-term funding alternatives to all-in cost and rate paths, and finally to two practical cases. The goal is to turn a rate view into a repeatable funding decision.

SOFR is the secured overnight financing rate, which is the cost of borrowing cash overnight against U.S. Treasury collateral in the repo market.

The New York Fed calculates it from actual Treasury repo transactions as a volume-weighted median and publishes it at about 8 a.m.

Eastern each business day using the prior day’s transactions. So it is a transaction-based overnight rate, not a survey.

And that word overnight matters quite a bit, because a single SOFR print can jump, while a SOFR index advance experiences the full rate path over its term.

On this slide, we will take a look at four drivers of movement in SOFR.

First, Fed policy and the front end of the curve.

SOFR is not the Fed funds rate, but they both live in the overnight complex.

The federal funds target range in administered rates like IORB help anchor the front end, so when the Fed hikes or cuts, SOFR generally moves with it. Second, repo markets supply and demand.

If there is more treasury collateral that needs to be financed or less cash in dealer balance sheet capacity available to finance it, repo rates can move higher.

More available cash and balance sheet capacity generally work the other way. Third, technical dates.

Around month-end, quarter-end, and year-end especially, institutions can engage in what is often called window dressing.

Increasing borrowings to improve liquidity ratios around reporting dates, driving up demand and SOFR rates and spreads.

And fourth, market stress and balance sheet pressure.

September 2019 is the classic example of a true repo market dislocation.

SOFR was 2.43% on September 16, then jumped to 5.25% the next day as corporate tax payments, treasury settlements, and title reserve and balance sheet conditions collided.

But March 2023 gives us an equally important counter example.

When Silicon Valley Bank and Signature Bank failed, demand for liquidity across the banking system surged.

Total FHL Bank borrowing ultimately reached about $804 billion in the first quarter, up 37% from the prior quarter.

Yet SOFR itself barely reacted to the initial banking stress.

It remained around 4.55% through the first several days surrounding the failures. That distinction matters.

A banking liquidity event does not automatically become a SOFR event.

For SOFR to materially dislocate, the pressure generally has to reach the Treasury repo market itself.

On this slide, we will take a look and see what these movements in SOFR look like visualized.

On the chart on this slide, the green line is daily SOFR and the shaded band is the federal funds target range.

The first thing that stands out is the step pattern.

When the Fed changes the target range, SOFR moves with the front end almost immediately.

Between those policy moves, you can see smaller fluctuations from the repo market forces we just discussed. That is an important distinction for a borrower.

Most of SOFR’s movement is the systemic rate cycle, Fed hikes and cuts moving the overnight complex.

Then, layered on top of that, you can get smaller, technical moves from repo conditions, and occasionally, as in 2019, you can get a genuine dislocation.

But 2023 reminds us that even severe stress somewhere else in the banking system does not necessarily mean that that relationship breaks down.

So when evaluating floating rate funding, the question is not simply, can SOFR go up?

Of course it can. The better question is, what would make it move?

How far could it and how long would that move have to persist before it overwhelms the starting advantage of floating?

And that framework brings us to the second part of the fear.

Even if SOFR does spike, what does that actually do to the cost of borrowing? The key is averaging.

The approximate impact on average cost is the size of the rate shock multiplied by the number of days it affects the borrowing, then divided by the number of days in the term. Take the first example.

A 50-basis point hike on day 26 of a 31-day month adds about 9.7 basis points to the average cost for that month, not 50 basis points.

Now look at the second example.

A 300-basis point spike lasting one day in a 90-day term adds only about 3.3 basis points to the average cost.

That example is intentionally extreme.

It is roughly the scale of the September 2019 SOFR shock, when the rate jumped from 2.43% to 5.25% in a single day. And importantly, the worst of that move did not persist.

SOFR was back to 2.55% the following day. So the takeaway is not that spikes do not matter.

They absolutely do.

The takeaway is that the economic impact depends on the magnitude times duration. A headline print tells you how high SOFR got.

The average tells you what you paid. Now we will match structure to the funding job.

Daily cash manager or DCM is for true overnight or uncertain funding.

It provides you maximum flexibility with daily repricing and no term certainty.

A SOFR index advance gives term liquidity with a floating rate that can fit a recurring or seasonal balance that is more durable than a daily swing without automatically paying for fixed certainty.

A classic advance fixes the rate for the selected term, and that is useful when the need is known and budget certainty matters.

But market expectations are embedded in the quote, which is something we’ll explore a little bit more in a few slides when we compare a one-year classic to a one-year SOFR index advance. Payment timing matters as well.

For SOFR index advances, interest payments are generally annually or at maturity, whereas for classic advances, they’re usually monthly for interest payments.

However, for classic advances under six months, there is the option for a bullet or at-maturity interest payment.

So what we’re looking at is not three different products or rates, it’s three different solutions to three separate balance sheet challenges.

To make this comparison concrete, we’re using actual FHL Bank Boston and market pricing.

In the rate snapshot underlying the analysis, SOFR was 3.57% and the one-to-three-month SOFR indexed advance started at 3.75%.

For reference, these rates are from July 20, 2026.

That compares with 3.91% for a one-month classic, 4% for a three-month, 4.08% for a six-month, and 4.18% for the one-year Classic, as well as 3.93% for DCM.

What matters here is the starting relationship.

Floating begins below the comparable fixed alternatives, creating a cushion that must be exhausted before rates move against us enough to change the answer.

Now the question becomes, does that day one advantage survive an all-in comparison? DCM requires an activity stock purchase equal to three percent of the advance.

Non-overnight advances require a 4% activity stock purchase.

At the 6.71% dividend assumption shown here, which is the Q1 dividend from FHL Bank Boston, that is an estimated 20.1 basis point benefit for DCM and 26.8 basis points for term advances.

Estimated all-in cost becomes 3.73% for DCM, 3.48% for one-month SOFR and 3.64% for the one-month Classic.

So under static rates, SOFR retains about 16 basis points of cushion versus Classic and 25 versus DCM.

Now we stress it.

Here is the dynamic test.

We are comparing the full-term cost of a one-year Classic to a one-year SOFR-Indexed Advance under five Fed Fund scenarios.

Here we are back to comparing sticker prices opposed to all-in post-dividend cost as we did on the last slide.

The one-year classic stays at 4.18% over the term.

The modeled SOFR index cost runs from 3.49% with two evenly spaced cuts to 3.78% with no moves from the Fed to 4.07 % even with two evenly spaced 25 basis point hikes.

As of this recording, futures price roughly one to two hikes over the In our one-hike scenario, floating is 3.93% versus 4.18% fixed, and even at two hikes, floating is lower at 4.07%.

That is the timely piece.

The evergreen question is the breakeven.

How much do rates have to rise, how early, and for how long before the premium for certainty pays for itself?

Now we move from product comparison to funding strategy.

Here at FHA Bank Boston, we see balances rolled in DCM or one-month classics month after month, sometimes for years.

At some point ask, are we really saying the whole funding relationship is overnight?

If a balance keeps returning, separate the base from the volatility. Keep true daily uncertainty in DCM.

Use shorter SOFR terms for less certain recurring needs, and then use three-, six-, or 12-month SOFR for portions that have proven more stable.

That is what the latter shows. Every maturity is a decision point.

If deposits grow, let a layer roll off. If the need remains, replace it.

If it proves durable, extend it. Terming out does not mean marrying the advance forever.

Callable SOFR adds another layer of optionality and for cheap too.

For the extra spread embedded in the callable quote the member can select cancellation dates and repay without a prepayment fee at those points subject to the terms.

That can be a way to have your cake and eat it too.

Term liquidity today with a planned exit if organic growth or runoff changes the need.

Now let’s put dollars around our initial thesis which was if it keeps rolling it is already floating.

For a 25-million-dollar funding layer under a scenario of two 25 basis point hikes rolling DCM costs about 1.05 million dollars.

Rolling one-month Classic is about 1.031 million and rolling one-month SOFR about $996,000 that is roughly $58,000 of savings versus DCM and $35,000 versus the rolling Classic.

All three carry short-term funding beta. They simply realize it differently.

DCM reprices daily. SOFR accrues through the path.

A one-month classic delays repricing until maturity, then the borrowing resets.

So fixed for 30 days is not fixed economically over a year of repeated rolls.

The second case makes the tougher.

We compare a one-year classic with a one-year Callable SOFR Advance with the option to call penalty-free at six months using the all-in post-dividend rate. The key difference from slide 11 is timing.

There, two hikes were evenly spaced.

Here, they are front-loaded into August and September, so floating spends much more of the year at the higher rate.

The chart shows the estimated all-in path; the table translates the scenarios into nominal interest dollars.

With no hike, Callable SOFR saves about $91 ,000 on $25 million of funding.

With one hike at six months, it still saves about $60 ,000.

And with two early hikes, the result flips, and the Classic is about $17 ,000 cheaper.

And that is why this scenario belongs here.

shows where the breakeven lies.

And the two early hikes that we use here are tougher than what the market is currently pricing in, and that’s intentional. This is the adverse path.

Also, the six-month cancellation date has option value here.

If deposits arrive or the need changes, the member can reduce or exit without a prepayment fee at that scheduled date subject to the So, floating does not win every path.

The value is knowing which path makes it lose. The takeaway is not that SOFR always wins.

It is that a recurring short-term role should not become an automatic habit.

Let’s follow the slide from top to bottom. First, let’s define the need.

Is it overnight, weekly, monthly, seasonal, or durable? Then, separate the recurring base from true volatility.

Compare the quoted rate and all in economics.

Activity stock, dividend assumptions, and payment timing should all be included here.

Then, let’s run the path.

SOFR depends on the average rate over the term, while maturity and reset timing determine how much of that path each structure lives through.

Finally, match the structure, DCM for volatility, short SOFR for the near-term base, and longer SOFR, Callable SOFR, or fixed funding as the need becomes more durable.

That turns what does the Fed do next into the better ALCO question, which is what risks are we carrying, what are we being paid to carry it, and where is the breakeven? And this is the analysis that we can run with you.

FHL Bank Boston has purpose-built tools that can use your actual borrowing amount, current alternatives, stock and dividend assumptions, payment and reset schedules, callable features, and different rate paths.

We can turn all that into estimated all-in rates, interest dollars, present value effects, break evens, and layered funding options for ALCO.

So, if you have a balance that keeps rolling, send us the pattern and assumptions.

We can model the alternatives your numbers.

Thank you for your time and for considering FHL Bank Boston as part of your funding strategy. Make the next role a decision, not a habit.