jeffyscott wrote: ↑
Sun May 06, 2018 12:10 pm
I am not sure how the proposed alternative of brokered CDs function, but assuming there are no additional costs and 2.9% is available, then break-even inflation vs. I-bonds would be 2.6%.
I just got a smidge over 3% net on a secondary 3-year CD on Friday. There is a commission of $1/$1000 face value for secondary market, but that is factored into the net yield. There is no cost for the new-issue 3-year CD at 2.9%, but 3% after commission is better than 2.9% with no commission. So let's call breakeven relative to I Bonds 2.7%.
However, if I'm looking at a 3-year holding period, I'd probably favor TIPS over I Bonds, as my estimate of the 3-year TIPS expected return is about 0.6% if held to maturity, so that's 30 basis points annualized higher than TIPS, or we could say that it's about twice the expected real return of an I Bond. So to be fair, BEI of the CD at 3% is more like 2.4%.
To argue with myself a bit, you only earn 15 bps per year of extra yield for extending maturity to three years with the TIPS from about one year with the I Bond (you can't sell the I Bond for one year, and then pay the 3-month penalty if you sell before five years). That's OK, but it isn't particularly steep, so maybe I would first go with annual allotment of I Bonds, then use TIPS for more meaningful portfolio inflation protection.
And this leads to a final point, that what I'm doing with CDs is not really a "proposed alternative", since I couldn't do with I Bonds what I'm doing with CDs. The annual purchase limits and use only with taxable-account proceeds preclude it. I could take a very small portion of the taxable cash I have been (and will be doing more) investing in AA/AA munis and to a lesser extent nominal Treasuries, and put that into I Bonds. Maybe I'll do that--it's not either or.
If I add in the effect of my 7.5% marginal state tax rate, then it's more like 2.4%. With the brokered CDs in tax advantaged space, the break-even point would be somewhere between those, as they would displace something else and eventually be taxed.
I am generally buying CDs in tax-advantaged, not taxable, so the state tax exemption doesn't come into play. In taxable I would compare to nominal Treasuries (same state tax exemption) or AAA/AA munis, and I would look at taxable-equivalent yields (TEYs). For me, TEY of a 3-year Treasury at 2.63% is about 2.95% (after state tax exemption factored in), but I probably would beat that with an AA or maybe even AAA muni (historical 3-year default rate for AA is 0.00%). In mid-April I bought a 21-month AA muni at 3.1% TEY.
Not sure how "displacing something else" factors into the yield calculations. For me they wouldn't displace something else, because I would be using proceeds from maturing 5-year direct CDs, transferred from bank or CU into Fidelity. So I would be deploying this cash into TIPS instead of CDs (and threads like this motivate me to look more closely at adding some TIPS into the mix). But of course for you it is different.
I don't think of my traditional IRA as "eventually being taxed", but as split into my share and the government's share based on whatever tax rate I end up paying on distributions, so my share is tax free. Or if it's easier to think about it, just consider that it's in a Roth IRA (if you properly account for taxes, traditional and Roth are essentially equivalent). My most recently matured CD was indeed a Roth IRA CD at PenFed, and the proceeds are now in my Fidelity Roth IRA, being deployed into CDs at this point.
If I use the St. Louis Fed expected inflation figure of 2.23%, the "insurance cost" is about 0.27% +/- 0.1%.
I prefer to use market-based BEI rather than "expert" forecasts, and for any of these we have to look at the time period for the forecast. Is it one year, five years, 10 years? If I'm buying a 3-year CD, I'm interested in the market's expectation of inflation over next three years, and will use that as the expected value, knowing that there's a dispersion of probabilities around this expected value.
At any rate, with 3-year real yield at about 0.6% and 3-year Treasury at about 2.6%, expected 3-year inflation is about 2% (maybe closer to 2.1% if I use more precise values for 3-year TIPS and Treasury). So a 3-year Treasury in IRA or 3-year AA muni in taxable at 3% TEY gives me about 90 basis points of inflation protection relative to market expectations. This is about 30 bps better than using Treasuries in an IRA (or for institutional investors where the state tax exemption probably isn't a factor, and who are setting the prices), so I'm essentially getting 30 bps of unexpected
-inflation protection with the securities I'm buying.
For our personal situation, high inflation is probably our biggest risk as pension is not indexed to CPI. Apparently, I'll be paying $27 per year for this allotment of inflation insurance via $10K in I-bonds to be purchased this month.
Yeah, when you look at dollar values for small amounts and low yields, you can almost always make an argument to make whatever point you want, like 2% isn't much better than 1% if you're investing $1,000 (or even $10,000). I've seen a similar point made multiple times just in the last week or two. Same argument can be made for expense ratios, yet most Bogleheads seem to be pretty adamant about keeping ERs as low as possible.
But sure, you're certainly are not taking much risk with $10K in I Bonds, but neither are you adding much inflation protection for anything other than a small portfolio. No harm, no foul (or perhaps, not much benefit, no home run).
I-bonds do have the added advantage of not displacing anything else from tax-advantaged accounts. We have essentially all assets invested in tax advantaged accounts and I-bonds, except about 1.5% of assets in a few savings and checking accounts. I'm working on reducing that 1.5% by getting rid of the savings accounts in favor of I-bonds.
But what would you displace from tax-advantaged accounts? If inflation is your main concern, then I assume you are holding mostly if not all TIPS or TIPS funds in your tax-advantaged accounts in preference to nominal bonds or bond funds. Correct? If not, then displacing a nominal bond or bond fund with TIPS seems like it would make sense for you. Again, no reason not to do both--buy I Bonds for low term-risk, low real return, and TIPS for more term risk and higher real return. As I mentioned earlier, from a liability-matching perspective, I Bonds only make sense for liabilities up to a year or two, so once past accumulation mode, they would be the first to be used, and pretty soon you'd end up owning only TIPS (unless you've been accumulating I Bonds for many years, in which case you also would have done better by buying longer-maturity TIPS, possibly even in a taxable account).