Post #547: The flip side of your declining 401K

Posted on March 11, 2020

The actual 2020 figure is nominally $34.5M.  But it’s not quite as simple as that.

This is just a quick post to calculate what interest rate the Town paid on its (nominally) $34.5M bond issue sold yesterday.  The results of that bond sale are reported on the Town’s website at this link.

The Town agreed to issue bonds with a $34.5M face value, a 1.86% nominal interest rate, and a $3.1M premium.  That last bit — the premium — is the confounding factor.  That extra money is the reason that the true interest rate isn’t 1.86%.  And neither the Town nor any other social-media-type discussion that I have seen has managed to explain it correctly.

So let me explain exactly what that is. (And, separately, in a different post, I’m going to try to track the premium dollars from prior bond issues, because they effectively are not reported with the Town’s capital accounts.  I tried but failed to do that last year, with the 2018 premium.  Those dollars have to be reported somewhere, I just have to find them.)

What is a premium?  It’s a payment above the face value of the bonds.  Investors aren’t going to pay $34.5M for these bonds.  They will, in fact, pay ($34.5M + $3.1M =) $37.6M.

Somewhat oddly, while the Town only needs $34.5M to fund its capital projects, the Town is actually going to collect $37.6M.  So the Town is actually collecting more money than it needs to fund those capital projects.  It’s collecting more than the $34.5M face value of the bonds.  So it’s going to end up with $3.1M that hasn’t been “spoken for”.

Even more strangely, the Town only repays the face value of the bonds.  That is, it only repays the $34.5M plus interest.

That makes it sound like the Town is getting free money, doesn’t it?  But that’s not true.  The iron rule of market transactions is that nothing is free.  And in this case, the interest rate adjusts to pay the investors for their premium.

So this is free money, only in the sense that you get a “free phone” with a three-year phone contract.  You pay for it in the monthly (or in the case of bonds, annual?) payments that are higher than they would be in the absence of that “free” phone (or that bond premium).

(Points on a mortgage loan illustrate the same type of tradeoff, but in a direction opposite of a bond premium.  With points, the borrower pays additional money up front, then pays a lower interest rate over the life of the mortgage (than they would absent the points.)  With a bond  premium, the borrower receives additional money up front, and pays a higher interest rate over the life of the loan (than they would absent the premium).

The point of doing it this complicated way — with a premium — is to avoid taxes.  (Or, more properly, avoid the potential for maybe having to pay taxes on some of the income from these bonds, depending on future market conditions).  That’s an extras-for-experts that is explained  on this website.  (Note that I did not say “easily” or “clearly” with that.  Explained, yes.  Easily, no.)

After all, the advantage of municipal bonds, from the investor’s standpoint, is that they are tax-free.  But only the interest payment is guaranteed tax-free.  Capital gains on those bond values are taxable, and (as explained at the link above) the point of setting the bonds up this way — with a premium — is to minimize the risk that someone will have to take capital gains on these bonds as ordinary income (rather than at the capital gains rate) under the IRS “de minimus” rule.

OK, with that squared away, what were AAA 20-year muni bonds yielding yesterday, on average?  (I should have gone looking for this information yesterday, as interest rates popped up quite a bit today).

The only average yield estimate I can find for 20-year top-rated muni bonds is shown below, for today.  FSM shows 20-year AAA muni bond averaging just 1.25% yield, as of today.  Presumably, that would have been somewhat lower yesterday.  (Data from Bloomberg show a 0.24 percentage point increase in the yield on 30-year muni bonds today, consistent with interest rates moving away from the panic levels of earlier this week.)

Source:  Screenshot from FSM.

No matter how you slice it, it appears that the timing of the bond issue was nearly perfect.   You can see that from Bloomberg’s graph of the 30year rate.  Yesterday was a good day to issue a bond.

Source:  Bloomberg.

But from either source of interest rate data, we’d have expected a much lower interest rate than the 1.86% cited by the Town of Vienna.  Putting aside various un-quantifiable differences (bonds vary based on who issues them, and so on), does this mean that Vienna overpaid for its bonds?

No, because of the premium.  The 1.86% is only paid on the face value of the bond.  The actual yield is lower than the 1.86% interest rate on the face value, because investors paid more than the face value for those bonds.  That’s the premium.

(Think of it this way — the Town could use those premium dollars to pay off part of each bond payment.  So the Town could certainly structure this so that it’s paying, on net, less than 1.86%.  The only trick here is to say exactly what the Town is paying.)

Now I can do what I set out do to, which is to calculate the interest rate (yield) that the Town is actually paying.  What’s the effective interest rate, from the Town’s stream of payments, for an initial total initial investment of $37.6M.  That uses two Excel functions.  The first is =PMT(0.0186,20,34500000), which shows that if the town pays this back in equal annual installments, those payments will be about $2M per year.  And then NPV, adjusting the interest rate so that those payments yield a net present value of (rounded) $37.6M.

To the contrary, the large premium means that the Town actually paid an interest rate of just under  1%.

Just to spot-check that, with interest rates this low, compounding hardly matters, and you can practically reduce this to simple interest.  So you can check the math with simple arithmetic.  Over 20 years, the Town will pay out (20 x the figure in red) = $41.6M.  So, in total, it pays back just $4M more than the original receipts of the loan ($41.6M – $37.6M).  So, per year, over 20 years, the “true” interest portion of payment is just about $200K per year.  (True, meaning, figured on an initial reveue of $37.6M).  If I pretend that this is simple interest (not compound), and the balance was paid down linearly (so for the average year, you only owe half the money), that would amount to simple interest of 1.1%/year.  So, yes, an estimated compound interest rate of just under 1%/year is plausible.

Given that the trailing rate of inflation is about 2.5%, the Town is paying a real interest rate about -1.5% (negative one-point-five percent).  In the current climate, investors are giving money away, more or less.  And the Town took full advantage of that.

Does this mean that borrowing such a large sum was prudent?  Not necessarily.  Most of what gets paid back, as noted above, is the principal.  In the last recession (2008), the revenue stream that pays for these bonds (meals tax) was hard-hit.  And this large borrowing assumed significant sustained growth in meals taxes to cover the repayments.

The same factor that is resulting in flight-to-safety panic-level interest rates — coronavirus — is likely to impact willingness to (e.g.) eat in public places like restaurants.   So, while we benefited from the interest rates, don’t lose sight of the fact that the same factor that drove down the interest rate is likely to have a negative impact on meals tax revenues.  We’ll just have to see how that all works out.

Addendum:  Sure enough, USA Today’s article on the coronavirus 3/12/2020 has a claim that one-third of Americans plan to dine out less.  Based on some data source.  Whatever “less” means.  So that’s the extent of hard data on this topic, for now.