FREE NEWSLETTER
BILLIONS OF DOLLARS poured into Series I savings bonds toward the end of October, as investors rushed to snag the 9.62% annualized rate then on offer, which was guaranteed for the first six months. But it turns out these folks were a tad too hasty.
How so? Buyers of I bonds are promised a pretax return equal to the inflation rate, plus they sometimes also get an additional fixed rate of interest, over and above inflation, depending on when they buy. For the past two-and-a-half years, that additional fixed rate of interest has been zero. Pretty much everybody—including me—assumed it would remain that way. After all, with inflation so high and with billions flooding into I bonds, why offer anything more than a fat inflation-driven yield?
But it turns out those sneaky folks at the Treasury Department had other ideas. For the I bonds sold during the six months starting Nov. 1, the annual fixed rate has jumped from zero to 0.4%. One possibility: Perhaps the Treasury Department did this because Treasury Inflation-Protected Securities, or TIPS, are now also offering higher real yields.
The result is that, for the first six months that today’s buyers own their I bonds, they’ll earn an annualized 6.89%. But what’s really guaranteed is 3.44% for six months, or 3.24% to compensate for recent inflation plus half of the 0.4% fixed rate. Thereafter, today’s I bond buyers will get a return equal to the inflation rate, plus 0.4% a year.
What if, instead, you’d bought in October? You would pocket an annualized 9.62% for the first six months, equal to 4.81% for that six-month period. That’s better—1.37 percentage points better, to be precise—than the 3.44% that November’s buyers will collect during their first six months.
But after the initial six months are over, things start to change. October’s buyers will get a yield equal to the inflation rate, while November’s buyers will get inflation plus 0.4% a year. It won’t take many years for today’s buyers to catch up with October’s buyers, and thereafter they’ll pull further and further ahead.
All this carries something of a sting. Why? You’re limited to buying $10,000 of I bonds per year, plus another $5,000 using your federal tax refund, assuming you owe that much. On top of that, you can’t sell savings bonds in the first 12 months and you lose the last three months of interest if you bail out in the first five years. Still, October’s remorseful buyers will get another chance in 2023—when they can invest $10,000 more.
I am not certain if this article is more helpful or more harmful. The comments suggest that people are engaged with the topic and therefore it is helpful. However, is it helpful or harmful to point out (in retrospect) that people would have been better waiting? Does it reinforce the notion of an optimal investment strategy that once again we failed to achieve (even though “optimal” could only be known after the fact)? Based on the comments, it seems like the calculations needed to understand the impact of the decision are complex and multiple intelligent people are coming up with different answers to the question. Is the impact of the sub-optimal decision dwarfed by the bigger accomplishments of saving enough to have $10,000 to invest and all of the benefits of Series I savings bonds that have been discussed in previous articles on Humble Dollar? One of the other comments links to this topic on the Boggleheads website. Over there, this topic makes sense. On Humble Dollar, I am not so sure.
Why is the article helpful? It seems many recent buyers of I bonds were attracted by the high nominal yield — guaranteed only for six months — while not fully understanding that they were buying an investment that would merely track inflation and, after taxes, lose ground to it. Indeed, even as billions poured into I bonds in late October, these buyers could instead have bought Treasury Inflation-Protected Securities in their IRA or 401(k) and earned a return with a healthy yield over and above inflation. November’s I bond rate announcement offered a chance to explain how I bonds work — and how buyers need to look beyond the high initial yield and think about what their long-term return will be.
Don’t tell me that you’re evaluating a transaction after the fact! If someone’s plan – without firm knowledge of the future change – was to buy the bonds in October, it’s hard to fault that. Market timing is imprecise!
A seemingly sophisticated break-even analysis was published elsewhere that assumed a 0.5% fixed rate ( https://www.bogleheads.org/forum/viewtopic.php?p=6921776#p6921776) and estimates a break-even of 7.5 years (ignoring the tax drag of selling earlier than otherwise). The 0.4% fixed rate earns you an additional $40 for the first year for a $10,000 I-bond (and $40+ yearly thereafter due to compounding), while buying in October yields an initial $481 (the interest on which is compounded) that you’d never get with a November purchase but forgoes the last 6-month inflation rate adjustment that the November bond will get. So I think the answer depends on an unknown future inflation adjustment many years from now (but will hopefully be a lot less than October’s adjustment rate). I decided to split the difference: I bought a gift I-bond for my wife in October to be delivered in January, and will buy a new I-bond for me in January. My plan is to sell the October bond whenever it makes sense to do so, preferably after 5 years and before the break-even point.
Bought in to I bonds ETF TIPS with 40k (20k for me, 20k for wife) in August knowing that fund expenses would cause a somewhat lower return but felt it worthwhile to avoid the usual government rules & regs and with the ability to sell when desired without penalty. Also noted that I bought in near the fund’s 52 week low at the time. But it’s been on the decline ever since, with a collective principal loss on paper of $3,242.00 today 11/5/22. Bought in just in time to collect the Sept 1st distribution of just under $500 collectively (7.55%), to which I thought wow at the time if this goes on every month. It hasn’t. Similar distributions were made for 6/1, 7/1, 8/1, but zero since 9/1. Nothing for October and apparently nothing now for November either. Buyer beware, as usual. Not sure what to think at this point and will be discussing it with my Fidelity Advisor next visit.
I was wondering if you would calculate the break-even point and publish the result and or any potential strategies for maximizing the investment. For instance, assuming I invested $10,000 in October would it make sense for me to buy I bonds in January 2023 and then sell my October bonds before the benefit of the higher overall October rate has been eroded?
It’s a trickier question than you might imagine. First, if you sell 2022’s I bond purchase, you will never again have 2022’s $10,000 I bond purchase. Are you happy to give up that annual tranche of I bonds? Second, you can’t sell I bonds in the first 12 months, so we don’t know what the rate on offer will be when you turn around and plow the sale proceeds back into I bonds. There are also additional complications, caused by the different yield adjustment schedule that October’s I bonds and November’s I bonds are on. But in the end, it seems the big issue is the loss of the last three months’ interest if you sell in the first five years. Because of that, selling in the first five years seems like the wrong move — and holding on for even longer may make sense, depending on your tax situation and what fixed rate is on offer when you hope to sell and then reinvest the proceeds.
Interesting. We had already bought our $10k each for 2022 earlier in there year. In October, we bought each other $10k in gift bonds, planning to deliver the gifts in 2023.
With this new info, we’ll have to think about whether we deliver the gift bonds as planned, or instead buy ourselves our $10k each in 2023, and hold onto the gifts in our gift boxes until it seems advantageous to deliver them rather than buy for ourselves directly.
I was also surprised by the fixed rate. The best it’s been since May 2019, and going back to the Great Recession, 0% has been pretty common.
I was one of many who was tempted by the 9.62% annual rate (all variable, 0% fixed) and purchased 10K under my social at that rate. I have no regrets, but thought purchasing more under my wife’s social might be better if I waited for a higher fixed rate. I’m glad this seems to have paid off.
I may gradually sell some of my initial 10K purchase in the next few years whenever the variable rate falls far enough and a fixed rate is on offer. That way the three-month penalty won’t be much.