Other than taxation issues, pretty much everything here is bikeshedding.
Had not heard that term before, so had to Google it: Law of triviality - Wikipedia
Parkinson's law of triviality is C. Northcote Parkinson's 1957 argument that members of an organisation give disproportionate weight to trivial issues. He provides the example of a fictional committee whose job was to approve the plans for a nuclear power plant spending the majority of its time on discussions about relatively minor but easy-to-grasp issues, such as what materials to use for the staff bike-shed, while neglecting the proposed design of the plant itself, which is far more important but also a far more difficult and complex task.
matto wrote:Keep in mind CDs have issues, if you buy above par they are not risk free.
Of course this only applies to brokered CDs bought at a premium on secondary market, and for those it is indeed something to consider. This is why I noted in an earlier reply that the secondary market 10-year CDs I found at about 3% were priced at a discount, so no exposure to no FDIC coverage of the premium amount.
I actually experienced this first hand in a small way when a Puerto Rican bank I bought a brokered CD from at a premium went bankrupt, and I lost the premium amount. Fortunately it was a small loss, but it implanted this lesson firmly in my mind.
Again, not applicable to direct CDs, new-issue brokered CDs, or secondary CDs bought at or below par (100).
Also if you buy a 10year CD at 5% and in 2 years the bank goes bankrupt, FDIC will *not* match the interest rate you had previously.
Yes, there is a small reinvestment risk with CDs. The historical default rate on banks is very low, so it's a small risk, especially if you stick with higher-rated banks and CUs.
This adds a small but non-zero variance to the left side of the probability distribution of expected return, while the possibility of doing an early withdrawal and reinvesting at a higher rate (that more than compensates for the early withdrawal penalty) adds some variance to the right side of the probability distribution for direct CDs (but of course not for brokered CDs). In my view, and experience, the probability distribution is skewed more to the right by the latter than it is to the left by the former for direct CDs. Of course for brokered CDs there is only the left skew due to reinvestment risk.
Note that for the left skew you must estimate the joint probability of bank/CU failure and the rate you can get on a new CD being lower than the current rate, while for the right skew you must estimate the joint probability of a sufficiently higher rate becoming available and the early withdrawal being allowed (I assume a small probability of an early withdrawal being disallowed, since CD terms usually have language that allows it).
Come to think of it, it's possible that you could reinvest at a higher
rate after bank failure, with no early withdrawal penalty (EWP), so there also is a bit of right skew for the bank-failure possibility as well, making the left-skew even less significant.
Finally, if you ever plan on rebalancing, note that the cost to CD rebalancing might be insurmountably high and you might be forced to hold to expiration.
First, this is a good reason to hold some Treasuries, other bonds, or cash, remembering that you probably only need a portion of your fixed income for rebalancing.
Second, the cost of rebalancing out of good direct CDs is unlikely to be "insurmountably high", with the early withdrawal penalty (EWP) limited to say 1-2%, depending on the rate and early withdrawal terms (most of my CDs have EWPs of six months of interest, so 1% on a 2% CD and 1.5% on a 3% CD). For brokered CDs, as long as the bid/ask spread is less than the yield premium, which I think it typically is, you should be OK as long as you hold your CD for more than one year.
Finally, rebalancing events, with say 5/25 bands, are relatively rare, so there's a decent chance that you can use proceeds from maturing CDs to rebalance into stocks, which is exactly what I did in early 2016. You can even incorporate maturing CDs into your rebalancing policy, which certainly is no worse than annual rebalancing. I use a combination of 5/25 and rebalancing with maturing CD proceeds, especially with CDs in taxable accounts.
I'll also add that holding to expiration makes you lose the term premium aka the roll yield, although you can compensate by going further out on your CDs/other bonds.
"Lose the term premium" is overstated. You get the term premium from higher yields simply by extending maturity, which is correctly implied by the second part of the quoted sentence.
Roll return is one component of the term premium, and it's an uncertain component at that. The expected roll return as of 11/8/2016 was completely wiped out in a single day on on 11/9/2016, and went quite negative relative to 11/8 through mid-December, as I illustrated in this post: What happened to my roll yield (aka roll down return)? - Bogleheads.org
For a five-year holding period, the yield premium of a good 5-year CD (compared to 5-year Treasury) is an almost certain thing, with the only uncertainty being bank/CU failure and inability to reinvest at the same or higher rate, while any roll return from selling a Treasury before maturity is highly uncertain. If you go with the philosophy of taking risk on the equity side, then you want more certainty in the returns of your fixed income, and potential roll return does not contribute to that certainty, but detracts from it.
However, to be fair, you get a right-skew from potential roll return in a steady or falling rate environment (assuming positive yield curve, especially if it's steep), while you get a right-skew from potential break/reinvest with a direct CD in a rising rate environment (or even just from a really good CD deal becoming available). The tie breaker for me is the yield premium of the CD, which gives the CD the win assuming both are held to maturity, so win/win/lose for direct CD vs. lose/lose/win for the Treasury, and win/lose/? for a brokered CD.