Choosing between a muni bond fund or an individual muni bond?

[Oilcans]

I am fairly new to bond investing and know most of the pro's and con's of each but when comparing the annual interest for each I am left confused.

For example FTABX for which I have $150,000 invested pays me around $405 each month. which computes to about 3.6% per year.

With the individual bond, CUSIP # 603827VS2 which I also own and paid $5,626.45 plus premium of $21.53 for a total of $5,647.98. For a 5% coupon rate I am paid $125 semi annually. This computes to around 4.4% per year.

Based on this wouldn't I only invest in the individual bonds since it is providing around 1% more per year? What am I missing?


Thanks for any advice.

[bluquark]

The SEC yield of FTABX is 2.08%. Muni bonds don't earn anywhere near even 3.6%, and haven't for a long time, so you must be making accounting mistakes somewhere, although I find it hard to fathom where the mistake could be. Did you make sure to only count yields, not capital appreciation, for that 405$ a month? Is the payment schedule slower than monthly?

In my view, it's almost always taking uncompensated risk to invest in individual munis. If you want exact duration control, use these instead: https://www.ishares.com/us/strategies/b ... nd-ladders

[bluquark]

For the individual Minnesota bond, https://www.municipalbonds.com/bonds/issue/603827VS2/ reports a "yield" of 1.430 for the last purchaser last week. I assume this probably means annualized percent yield in relation to purchase cost. It's a bit on the low side compared to the Fidelity bond fund so it's possible I in turn am making an error because I'm not used to looking at individual CUSIPs, but it's more plausible to see a too-low-to-be-good yield than a too-good-to-be-true one.

[Oilcans]

Thanks for your replies Bluquark:

I guess I am getting away from the "yields" and am computing my own "actual" numbers.

For FTABX I receive and reinvest $405 per month ($4,860 per year) divided by my investment of $150,000 = 3.24%.
For the individual bond I receive $250 each year divided by my investment $5,647.98 = 4.4%.

I also own an ETF fund (MUB) and am considering investing into more bonds. But based on my above calculations, almost all of my individual bonds outperform bond funds.

Maybe I am looking at this the wrong way??

[not4me]

I also own an ETF fund (MUB) and am considering investing into more bonds. But based on my above calculations, almost all of my individual bonds outperform bond funds.

Maybe I am looking at this the wrong way??

I didn't notice when you bought the individual bond, but obviously it has a set coupon rate. On the other hand, funds are continually buying/selling which affects how much they pay out. So, over time, if rates in the part of the bond market you are talking about are rising, the individual coupon rate may lag over time; if rates going down, the opposite occurs.

Additionally, individual bonds will keep lowering the duration & time until $ will need to be re-invested. Funds tend to maintain a steadier duration. This has the effect of absorbing interest rate changes more gradually than individual bonds.

Of course, there are other differences as well in looking at fees, diversification, etc that may or may not be important to your situation.

[hudson]

I also own an ETF fund (MUB) and am considering investing into more bonds. But based on my above calculations, almost all of my individual bonds outperform bond funds.
Maybe I am looking at this the wrong way??

Buying individual bonds is not for the beginner. I looked at buying individual bonds and gave up because of the complexity and the work involved. For me, it's so much easier to buy MUB, VTEB, or VWIUX...intermediate muni funds/ETFs. It's very easy to get a "haircut" when buying individual bonds.
When I was looking at buying my own, I couldn't find any AAA/AA bonds on the Vanguard or Fidelity websites that worked for me. I would just as soon let Vanguard's bond people do the work with VWIUX. I seriously doubt that they take any haircuts buying bonds.

[ofckrupke]

With the individual bond, CUSIP # 603827VS2 which I also own and paid $5,626.45 plus premium of $21.53 for a total of $5,647.98. For a 5% coupon rate I am paid $125 semi annually. This computes to around 4.4% per year.
[...]
What am I missing?

1) you paid $5626.45 for $5000 face value of this bond issue; therefore the premium you paid was $626.45. The $21.53 was the accrued interest you paid at acquisition to the seller for the partial coupon period before the transaction; this is about 31 days interest so you must have bought the bond about one month into a six month coupon period. You paid $112.529 per $100 face value; but since this bond recently struck for 115.244 and today's ask is 115.872, I think you probably acquired this holding around 6 or 12 or 18 (etc) months ago rather than yesterday.

Because this bond will repay at par (100) at maturity or upon being called before maturity, loss of the acquisition premium must be accounted in calculating its expected return. This bond matures on 1-1-2025 but is callable at/after 1-1-2024, and a call at the first opportunity would be the worst (non-default) case; so an appropriate yield calculation amortizes the premium over the period between acquisition and 1-1-2024 (if bought at end-Jan-2019 that would be 4 years and 11 months; if bought at end-July-2018 it would be 5 years and 5 months; etc). Since the path of constant yield used for premium amortization is close to a straight line for a 5% coupon over 4-6 years, it's both easy and useful to approximate the effect of amortization as a constant adjustment to your notional 4.4% coupon yield once the acquisition date is known - for example, supposing you acquired it at end Jan 2019 then correctly you would have anticipated that the premium amortization would eat about $127 of your $250 of coupon income each year, so your Yield-To-Worst at acquisition was probably about 2.2% rather than 4.4%. According to my broker's screen, today's buyer at $115.872 should expect YTW of 1.289% -- for today's buyer, this lower yield manifests as a higher premium, to be amortized over a shorter period, effectively taking a bigger bite out of the coupon interest rate.

2) the distribution yield on FTABX is higher today than its SEC yield. Roughly speaking, this tells you that that the yield curve over the range of maturities in its portfolio is, on average, lower than when its holdings were acquired. If the yield curve were to freeze going forward, then the fund's NAV would decrease over the course of time, recapitulating the amortization of an individual premium bond in the aggregate. Because of this effect, the SEC yield really is a more appropriate measure of a fund's expected return than its distribution yield (neglecting the effect of taxation anyway, which even for a muni fund can be nontrivial in the form of capital gain/loss when shares are sold). Since Fidelity charges 0.25% (net) to manage this fund, we can take 2.08% + 0.25 = 2.33% as the yield measure for its underlying bond portfolio.

Since the market prices bonds to reflect aggregate perception of risk, comparing the fund's portfolio's yield (ER-adjusted as above) against that of another bond or fund of like duration reveals their relative credit (default) risk; but in this case, the fund's duration of 6.76 years differs too much from the individual bond's 4.4y (to first call) for a useful comparison. This particular bond issue may be pre-refunded, in which case it would have negligible credit risk; but typically, the most important difference between FTABX and any individual municipal bond issue is the rlower idiosyncratic credit risk offered by the fund due to the breadth of its portfolio.

My advice is to learn the detailed math of yield on individual bonds and of funds as portfolios of bonds; and until you understand these much better than demonstrated in the OP, to stick with funds - for new acquisitions, anyway.

[bluquark]

2) the distribution yield on FTABX is higher today than its SEC yield. Roughly speaking, this tells you that that the yield curve over the range of maturities in its portfolio is, on average, lower than when its holdings were acquired. If the yield curve were to freeze going forward, then the fund's NAV would decrease over the course of time, recapitulating the amortization of an individual premium bond in the aggregate.

Thanks for the explanation, ofckrupke! I understand what's going on in my bond funds a little more concretely now.

What I'm still confused about is: wouldn't the bond fund apply the premium amortization before paying out yield to fundholders? How can it be that Oilcans got that 405$ put in their settlement account?

[ofckrupke]

2) the distribution yield on FTABX is higher today than its SEC yield. Roughly speaking, this tells you that that the yield curve over the range of maturities in its portfolio is, on average, lower than when its holdings were acquired. If the yield curve were to freeze going forward, then the fund's NAV would decrease over the course of time, recapitulating the amortization of an individual premium bond in the aggregate.

Thanks for the explanation, ofckrupke! I understand what's going on in my bond funds a little more concretely now.

What I'm still confused about is: wouldn't the bond fund apply the premium amortization before paying out yield to fundholders? How can it be that Oilcans got that 405$ put in their settlement account?

Thanks, you have identified the (underlined) part of my post that was most wrong.

To the extent that the fund acquires new issues offered with coupons above market rates thus settling initially with a premium, or buys bonds in the secondary market at market premium (as typical in an interim-declining yield environment), it is these sources of premium that the fund needs to amortize internally, thus distributing only the interest net of that adjustment. [In aggregate form, this (the fund homolog to amortization of individual bond premium) actually shows up in disparity between distribution yield and average coupon yield, not disparity between SEC and distribution yields.]

When the distribution yield exceeds the SEC yield, this reflects aggregate/average changes in yield environment post-acquisition of its holdings - none of which would be associated with amortization of premium accounting in an individual bond portfolio.

So, in the citation above I equated accounting of pre vs post acquisition yield environment variation...plainly wrong. I apologize for the mis-statement.

[not4me]

It might be worth pointing out for some who are learning this process that dealing with bonds purchased at a premium is different between bond funds and holding individual bonds. Expanding the scope a bit, this also differs when you talk about bonds that aren't tax exempt. Also, bonds in the fund might have been purchased at a discount or a premium.