My Approach to Bonds (and Cash)

I wrote about my overall asset allocation a while back (Part 1 & Part 2), but have been doing a fair amount of thinking late regarding the bond part, particularly in light of some more substantial cash savings from our remortgage that are eventually planned for home improvements/extensions, but with no clear timeline (at least years, realistically). Keeping my 10% in bonds plus more than another 15% of my overall investment portfolio in cash felt too conservative, but then do I just ignore the cash because it’s sort of short term savings? But those home improvement plans are ill-defined at best, and I hate to see that much money sitting on the sidelines for potentially years, at least without a coherent plan.

After a lot of back and forth, tinkering with asset allocations, playing with how to account for cash, reading about bond tents, etc., I came to two simple conclusions:

  1. Treat bonds and cash as a single asset category
  2. Increase my bonds/cash allocation to “my age minus 15” – making it 21% now

Bonds and Cash – one category?

We’re probably all aware of the challenges facing bonds right now:

  • Yields are miserably low – 1.31% on BND, 1.41% on BNDW, 1.04% on a 30 year gilt (0.13% on a 2 year! why bother?)
  • Interest rates have almost nowhere to go but up. I’m not predicting that rates will rise soon, since they’ve really been consistently low since the 2008 crash (small ramp up leading up to the pandemic, but still historically low). Maybe they move sideways for another decade or two, but, barring going substantially negative, they can’t go down much.
    • If interest rates stay flat, bond prices don’t change – all we get is the (very low) yield
    • If interest rates rise, the prices of bonds we own today go down
  • That results in what feels like a substantial downside (risk of prices falling) without a lot of upside.
UK (Bank of England) interest rates – US and other developed country ones look similar, sometimes worse!

But it’s worth reminding ourselves of the reason for having bonds at all, instead of just 100% equities:

  • Bonds are less volatile – even when their price changes, it does so slowly. Look at the chart below, specifically 2000, 2008, and 2020 where stock prices change rapidly and bond prices move slowly.
  • There might be an inverse relationship to stock prices, through the “flight to safety”. When stocks go down, some investors get scared and move into bonds. But again, look at the chart below – the bond price increases during stock crashes are far less than the stock decreases. If you wanted to make money during the crashes, you’d need to be massively tilted towards bonds, which would leave you a lot poorer in the good times. If your crystal ball is good enough to predict the crashes, you’d make a lot more money by shorting stocks than holding bonds.
  • For those of us with investments in multiple currencies, having bonds in our “home” currency can be a tool to manage exchange rate risk.
  • We know that equities tend to have higher growth rates than bonds, over the long term, so we are deliberately choosing less growth as a tradeoff for less volatility.
  • Some day, bonds may be useful to me as a stream of income, but right now I’d just as soon the money stay invested – I’m a good ways off before I start living off my investments.
  • Bottom line: bonds reduce volatility, at the cost of lower long-run returns. I want my bonds to preserve value more than I need them to grow – that’s the job of equities.
From the excellent Portfolio Visualizer tool

Putting those two things together, and bonds start to sound a lot like cash to me, in today’s environment: low yield (a bit higher than 1% is “good”) and want them to preserve value, at least in nominal terms. So why treat them as entirely separate asset categories? Now, this wouldn’t apply if you’re interested in particularly risky bonds – emerging markets, “junk” corporate bonds, etc. Those I would consider something different and closer to equities, and they’re not something I choose to include in my portfolio. But if your bonds are developed government or high-quality corporates, they feel pretty similar.

So, at least for me, it makes sense to mentally combine my bond and cash holdings into a single asset allocation.

What bonds & cash do I hold?

Most of the bonds I hold are pretty cash-like in the first place – the chances of losing value in nominal terms are remote:

  • Premium “Bonds” (59%): I put bonds in quotes, because I consider Premium Bonds cash rather than bonds. They’re guaranteed by the UK government and have no risk of loss, plus pay out at a 1% prize rate.
    • There’s some element of luck involved here, although I have tended to average around the prize rate – this is a large enough holding that the luck starts to average out. Aside from being one of the best interest rates available for cash today (you can get marginally better if you’re willing to lock up your money for a few years), they’re a bit of fun too!
  • TSP G Fund (22%): a unique holding for the US government employee Thrift Savings Plan, this pays a US government bond long-term interest rate (1.375% as of September 2021) but without risk of loss. The interest rate varies with overall US bond rates.
    • If I was able to move more money into the G fund, I’d increase this percentage and drop the Premium Bonds allocation, but since it’s basically a 401k for a former employer, there’s no way to do that.
    • I could reduce my exposure to other TSP funds (basically US and Developed ex-US equities) in favor of the G fund, but then I’d need to move money from Premium Bonds into those equities, which is a challenge due to all the normal hoops for a US citizen in the UK (PFICs, PRIIPs, etc.). The G fund is good, but not so good that I want to start a US & UK taxable individual US stock portfolio like I’ve done with UK stocks in our ISAs! Doing the US bookkeeping and taxes for 42 stocks in our ISAs is plenty; doing both US and UK bookkeeping and taxes for more stocks makes my head hurt. Plus I’m a believer in indexing, and am pretty confident that the diversification due to indexing is worth the 0.375% difference between G fund and Premium Bonds.
  • US Series EE bonds (10% of my cash/bonds allocation): these are gifts from my parents when I was younger, slowly maturing. Because they’re relatively old, the interest rates are relatively good, plus the guarantee for EE bonds to double after 20 years means a minimum interest rate of roughly 3%. I’m not buying new ones, but won’t be selling these until they reach maturity over the next 25 years.
    • I’m not buying new ones because I’m paying income tax on the accrual basis for these (a decision made for me long ago). That’s nice, in that I don’t owe chunks of tax when the bonds mature, but also means that any more EE bonds (or I bonds) would add to my tax burden while I’m in a relatively high US and UK tax bracket today.
  • UK gilt fund in my UK pension (<1%): I’ve just started building this position from £0, and it won’t amount to much for a long time (just a couple hundred pounds a month). But of all these options, it’s the only one exposed to the risk of rising interest rates, the only “traditional” bond fund. But it’s in a tax-deferred account, practically no risk of default, and starts to slowly build an exchange-rate buffer, since effectively all my expenses are in GBP but a lot of investments are in USD.
  • Other cash accounts (9%): I include our various checking/current and savings account in my bonds+cash number. This is small stuff at paltry interest rates (0.5% is “high yield” these days), just ready spending and very short term savings, a few thousand £ & $. I could just as easily exclude it, but my spreadsheet already has it calculated. Not planning on this growing over time, it might slowly shrink as I continue to whittle away at closing old accounts.

Other bond options and why I’m not using them

The options above are not terribly traditional, and I know not everybody has access to the G fund. A few thoughts on some of the other options:

  • Standard bond index fund (BND, BNDW, BNDX, or UCITS ETF equivalents): very low yields, more downside than upside in price. Simple comparison: you can get 1% in risk-free Premium Bonds or 1.3%ish in a bond fund with risk of the price going down if interest rates go up. Is that extra 0.3% really worth it? On £100k invested, that’s about a £3k difference over 10 years – not nothing, but not going to make a huge difference compared to the risk of loss.
    • You can tilt these however you like, in terms of risk, duration, etc., but it doesn’t change the basic problem. And if you get very risky to chase yield, this starts to feel more like the risk/benefit tradeoff of an equity, not a bond.
    • For me personally, this really doesn’t make sense compared to the G fund, at the same interest rate but no risk of loss.
  • Individual bonds (UK gilts or US treasuries): a bond ladder is an interesting idea, allowing tailoring of duration to meet your goals – once upon a time I had a CD ladder, back when CD interest rates weren’t rock bottom. A few challenges though:
    • Same or even lower yields as the bond funds, even at really long durations (which add to the risk of comparative loss if interest rates rise)
    • Tax efficiency: the US and UK tax bond interest at income tax rates, which means a lot of that low yield goes right back to a government. I could put them in my wife’s name to get around that, or in a tax-advantaged account. But that would mean my ISA for gilts, which has fairly horrific transaction fees compared to the return on bonds, or my Roth IRA for treasuries. In both cases, I don’t want to use tax-free (rather than tax-deferred) space for bonds, either.
    • Fair amount of effort for the yields. Between bookkeeping for US & UK taxes and the ongoing management, this is just more hassle than its worth to me.
  • Series I bonds: these are a pretty attractive option, and definitely worth considering. 3.54% interest (as of Sep 21), balanced with some lockup periods and restrictions on how much you can buy.
    • If it weren’t for the accrual basis that I’m already locked into using for tax on the interest, I would look at these more seriously. For those of you without that restriction, this could be a nice core holding, and you could defer tax on the interest well into the future, when you might be at a lower tax rate. For me, that would knock my after-tax return down to about the same level as the G fund (pre-tax, but easier to manage those taxes later on).
    • You do have to deal with the archaic TreasuryDirect website. It’s usable, but it’s straight out of the 90s.
    • More Series EE bonds would fall in a similar bucket. Some pros and cons compared to the Series I, I think I would tend towards Series I first, but also worth looking at.
  • Very high yield savings accounts: you can get 3%ish from a savings account, that sounds great right?! Wait, but you can only contribute £50 a month and only get that rate on up to £1,000? Sure, there’s a bit of money to be made here, but if you’re managing even a moderately sized portfolio this just doesn’t help much. No harm in doing it, but it winds up being a lot of effort if you want to get any significant amount of money into these kinds of accounts.
    • Bank account switching bonuses kind of fall in the same bucket. I’m not opposed to switching to make a quick £100, but that’s more of a side hustle than an asset allocation strategy.

Why 21% in bonds/cash?

First off, there’s no right answer to this question! Bogleheads have a great primer on asset allocation, if you haven’t read it already. But the gist is that there are plenty of rules of thumb, but that’s all they are. At the end of the day, you have to pick something you’re comfortable with, and “my age minus 15 in bonds” feels good to me. 21% is enough to provide some volatility dampening without having a huge impact on the return. It’s not nearly as conservative as “my age in bonds”, nor as aggressive as “100% in equities until retirement”. It’s somewhere in the middle, and it feels right to me.

It also puts me on a path to a bond tent, helping to address sequence of returns risk in early retirement. Some people would argue it’s early to start building that bond tent, but since 21% feels about right now, building another 1% a year over the next 15 years or so feels pretty reasonable. That would have me going into early retirement with a bond allocation around 36% – might play with that and bump it up a bit in the years closer to retirement, see how the world looks by then!

There’s a lot of “feels” in the description above – this is definitely one of the fuzzier bits of FIRE or investing generally. On one hand, you’ve got people saying that 50/50 is the logical default allocation, or that “your age in bonds” is plenty aggressive. On the other, you have people holding 100% equities all the way through accumulation and retirement. Without a crystal ball, there’s no way of knowing the optimal asset allocation, so all you can do is pick something that helps you sleep at night. I’ve been very fortunate to do well so far in my accumulation, and having a bit of bond/cash ballast in my portfolio helps me sleep.

My target asset allocation, as of September 2021

Other considerations – Inflation, Mortgage, & Emergency Fund

Some of you might be screaming by now “but what about inflation!?! You’re losing money in 21% of your portfolio!” And you’re right – inflation numbers are a bit murky with the effects of the pandemic, but it’s pretty clear inflation is higher than the 1 to 1.4% or so I’m getting from anything in my bond/cash allocation. And that’s ok – it’d be nice if interest rates beat inflation, but it’s hardly uncommon for them not to. The 79% of my portfolio in equities needs to beat inflation and deliver real returns – the bonds and cash don’t. Their job is to dampen volatility.

The obvious logical inconsistency is that I’m holding bonds/cash at 1 to 1.4% and paying a substantial mortgage at 1.26%. If you look at it a certain way, we took money out in our remortgage at 1.26% just to put almost all of it in Premium Bonds at 1%ish – that doesn’t make any sense! That was a deliberate decision though – we are paying a small premium (roughly £200 a year due to the delta in interest rates) in order to have the option of using that money for a future house extension. For us, that option is worth the price, and probably cheaper than it would have been to do a shorter remortgage, not take the money out, and then do another remortgage to access that money – not to mention the hassle or the risk that mortgage rates may rise in the next few years from their current extreme lows. So there’s a method to the madness, even if there’s also a very rational argument for not holding any bonds or cash that pay less than what our mortgage costs.

The last question might be “where does your emergency fund fit into this?” The answer is that we don’t have an emergency fund, at least not in the traditional sense. I’ll expand my thoughts on emergency funds in an upcoming post, but with 21% of our portfolio in cash/bonds, that’s about 2.5 years of spending in investments that are extremely safe, so I don’t see a reason to slap the “emergency fund” name on a separate pot of cash.

So What?

That’s a lot of words to describe how I’ve wrestled with today’s bond conundrum – and sadly, I don’t think there are many universal lessons, because this is a very personal decision. I just hope that my thought process might help prompt some thoughts of your own – I’d be very interested in hearing your views on asset allocation and how you are (or aren’t) using bonds in your portfolio, including any violent disagreements 🙂

3 thoughts on “My Approach to Bonds (and Cash)

  1. Another excellent post, thanks. I completely agree with you assessment of cash & bonds generally.

    I count myself very lucky that I married someone with a Defined Benefit (DB) pension! This pension (or the reduced survivor’s one) could be treated as a decent yielding index-linked bond portfolio, without the risk of capital loss. Of course, not all DB pensions are created equally, so there is still potentially a risk of scheme/employer insolvency (the Pension Protection Fund has some nasty income caps and is at risk of political interference). Unless your DB pension is backed by the government, but then they could still move the goal-posts.

    Following the work of Wade Pfau, if we didn’t have the DB pension, I think I would serious consider the use an annuity and/or a reverse mortgage in later life, but I don’t even want to think about the potential US/UK tax complications of these, so I would probably plan to renounce my US citizenship before going down this route.

    Also, I think once you’ve actually retired and have passed state pension age, you then could have the option of re-assessing your (by that point very high) bond allocation. Some argue if you’ve done outstandingly well and you don’t look likely to run out of money, that you could put all your money into safe assets. However, if you have children or support a charity, you may decide to stay invested in equities to potentially maximise any legacy – just so long as you’ve got yourself (and partner) covered first.

    I also treat cash/savings-accounts/government-bonds/emergency-funds as a single “safe asset” allocation. I used to have a UK index-linked gilt ladder, but I ended up making a killing when bond prices went up so I sold at a profit – I thought I should quit while ahead! I’ve occasionally considered the option of creating a ladder of “safer” retail corporate bonds listed on ORB. However, it doesn’t seem worth the effort when you could get similar rates from 1-5 year fixed-rate savings bonds – and as you say, you would likely end up paying tax at income tax rates.

    A UK friend, at least 10 years younger than me, recently asked what he could do with some spare cash for medium/long term investment. With all the usual caveats, I suggested he stick it all in Vanguard LifeStrategy 80% in an ISA (or SIPP if he was a higher-rate tax payer). Golly, I wish life was so simple for us US/UK tax residents…

    Like

    1. A DB pension would make a big difference! I’m not quite comfortable enough with Social Security or State Pension to really count on them as part of a bond portfolio – I do expect there to be something there, but who knows what.

      When I was looking at how the US taxes a UK pension, some of the publications also covered US tax on UK annuities (IRS Pub 939, in particular). I wasn’t paying especially close attention to those parts, but from what I remember and skimming over it again, there’s nothing that stands out as particularly punitive. Some complexity there, but it seems like the kind of thing that you have to figure out once and as long as it just keeps chugging away, the year to year part is pretty straightforward. It’s an 83 page pub but like 60 of those pages are actuarial tables, not TOO awful. And probably another one where you do a bunch of math just to prove that the US taxes are lower than UK and you don’t owe anything to the US anyway!

      I also like the idea of an annuity, tax implications aside. Biggest challenge today are the abysmal interest rates – it’s an awful lot of money to spend for certainty. You get the certainty that you’ll have enough to live on, but the equal certainty that all the money you spend on the annuity is gone forever. Tough one, will be interesting to see how it changes over the years.

      The post-retirement bond allocation is another interesting one. I totally buy the logic of a bond tent – shelter the early years from sequence of returns risk, but then return to the higher performing equities once you know you’ve survived that. But putting theory into practice is another matter, and one I expect is largely an emotional decision. Even now I can feel the competing pull of “maximize returns” on one side and “lock in your gains” on the other. 79/21 feels like a reasonable balance for today, but if I’m at 63/37 or so in retirement, will I really have the appetite to ramp up the risk? Especially if that risk is mostly for a legacy or charity?

      Good timing on the gilt ladder! It’s another one that sounds good in theory, but the practical effort and friction makes it unappetising. If I’m going to spend time doing bookkeeping, it’s more fun to do it on stocks!

      Of course, that’s the whole problem -I would very happily buy a Vanguard all-world equity fund and some kind of bond/cash thing and be done with it – ISA and pension, all set. Or same in the US, in a 401k and IRA. But the amount of hoops we go through to duplicate something so simple, it’s absurd.

      I think/hope I’m about at the point of the learning curve where my investments can be reasonably on autopilot for at least the next decade or so. Automatic investments in my workplace pension, transfer money to my IRAs once a year, regular savings and bonus into Premium Bonds and/or ISA, check asset allocation once a month and rebalance if it’s outside my tolerance bands. Taxes are getting close to stable too; my 2021 US taxes still have the legacy of some simplification (529 liquidation), but after that should be fairly stable. But still a lot more complex than it might be!

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: