Inflation-linked bonds

I guess most of you hold less bonds than stocks, if any. Same here. But bonds can be a great diversifier, especially for retirement/FIRE, which is why I will include them to my portfolio at some later stage of life.

I still can’t wrap my head around inflation-linked bonds though (e.g. TIPS for the US). People seem to be under the impression that these bonds protect their portfolio from inflation. But actually, imho, ILB only protect bonds from (unexpected) inflation, their overall return can still fall (rising rates situation / economy growth). ILB are still bonds, and their prices and yields move in opposite directions.

ILB typically perform well when future expectations of inflation increase, not when current measures of inflation rise. That means higher inflation expectations might potentially already be reflected in TIPS prices, and your TIPS might lose value.

It’s your real assets, such as stocks, that provide the best long-term protection against inflation.

So ILB are not the free inflation-protection lunch some people think they are, right?

1 Like

I view inflation-linked bonds as part of a long-term strategy. I agree they only protect against unexpected inflation. Now that we have inflation, it is a bit too late to jump in. But people who anticipated and bought them years ago are now better protected.

Personally I rely on stocks and real estate for inflation protection. The effects are not immediate but this is great because it means we get those at a discount now.

2 Likes

Thanks, that’s reassuring, exactly the way I see it.

So people who buy ILB essentially make a bet on rising inflation in the future?

I’m asking myself if I should permanently hold ILB as part of my portfolio, because I’ve already got stocks as inflation-hedge, and I don’t like placing “permanent bets” in my portfolio :grin:

Also, I assume nominal bonds will price in inflation too eventually?

And you raised an interesting point: ILB will not provide immediate inflation-protection, thus not protecting my purchasing power in times of an inflation shock. What might help here?

For locals, Government issued inflation linked bonds, have duration or interest risk, and that’s it. If held to maturity, none of those risks applies and they are as close to riskless as you can potentially be. The returns in real local currency you are promised while buying the bond are exactly the real returns you’ll get if you hold it to maturity, less taxes (and TIPS, for example, are not exactly tax efficient for people taxed in the US). This is independant of whether inflation is higher or lower than expected, it just is.

Locals can also use them to gamble on inflation. If inflation is higher than expected, they’ll outperform nominal bonds of the same duration. If inflation is lower than expected, nominal bonds of the same duration will outperform them. If their expected liabilities are in nominal terms (mortgage or car payments for example), buying TIPS would be a bet on inflation. For all other expenses, which coustitute the bigger part of what we are exposed to (groceries, utilities, entertainment, others), TIPS are a way to avoid betting on future inflation. For most use cases, the inflation bet is made by buying nominals to cover liabilities that are subject to inflation.

Inflation linked bonds funds are also some kind of a bet: fund managers usually don’t hold the bonds until maturity, so there’s some form duration/interest risk added into the mix. To avoid adding more risk, one should consider holding them for the duration of the fund, which is not always practical as our liabilities go closer.

None of this applies to us, as foreign investors. Inflation linked bonds protect from local inflation by adjusting to the CPI (or other similar mechanics, I don’t know British ILBs or other ILB than the US ones). The US (or other local) CPI isn’t the inflation we are subject to. Currency risk applies on top of that, making them volatile assets for a foreign investor, which is not usually what we are searching in bonds.

To me, the bottom line would be:

  • Unexpected liabilities should be covered by cash (emergency fund).

  • Bonds are for expected liabilities, inflation protected bonds are best to cover liabilities that are subject to inflation, which is most of them.

  • Buying bonds to temper portfolio volatility and zagging when stocks zig is a bet. In that case, I’d not limit myself to bonds and try to add additional uncorrelated assets (but this is a different discussion altogether).

3 Likes

Thanks, that’s a helpful summary!

What I don’t understand is the ILB mechanism: For TIPS, I believe, inflation-adjustment only happens every half year, retrospectively. That’s too late for most grocery needs, isn’t it?

And couldn’t you just aswell use nominal 6-months bonds for that purpose, which would be pricing in inflation by then (providing higher interest rates to maturity)?

Also, I believe the minimum duration of TIPS is 5 years. That makes them quite volatile to cover your immediate purchasing needs, doesn’t it? And stocks/nominal bonds might be a better investment at such a long duration anyway?

Interestingly, I Bonds as an immediate-term bond are only offered with very tight restrictions, you cannot hold more than 20k per year, you can only cash in (redeem) your I bond after 12 months. And, if you cash in the bond in less than 5 years, you lose the last 3 months of interest. I guess they’d be to bad of a deal for the US Gov. otherwise :grin:?

I don’t have a full understanding of it but when I read about liability matching, my understanding is that it is mostly the principal that is expected to cover the liability. I’ve yet to read something suggesting using the interests of TIPS to cover ongoing expenses. My guess is that that is because the amount of capital that should be tied in such a scheme would make it impractical.

So, when I write about liability matching using bonds, I’m thinking that the principal should come due when the expense does. In the context of a retirement portfolio, if my planned expenses for the next 6 months are $20K, I would want to have some bonds with principal value of $20K expiring just before that. You are right that the interests suffer from the 6 months differential, as do the principal from tax drag.

For short term liabilities, I don’t know of a short term instrument that would protect against inflation. I would guess that for short term or ongoing expenses, the model is either to be employed, and having a salary that adjusts for cost of life, or, as a retiree, to have already built a portfolio allowing to match liabilities as they come (for example a bond ladder).

The instruments I’d use to cover short term liabilities would be savings accounts, short term bonds/bills, or Certificates of Deposit (for US people) or Kassenobligationen (for Swiss residents with liabilities in 2+ years), which can also be used to build a ladder so that, past the first stages, some principal comes due regularly allowing to cover expected expenses, albeit without automatic adjustment for inflation.

My understanding is that I Bonds are a program offered by the US government to help alleviate the burden of inflation from their residents. I would see it as a form of subsidies. The limitations are probably there to limit its costs.

3 Likes

Yes, this makes perfect sense to me, thanks. And I think you’ve nailed it, I’ve never thought of IBonds as actually being subsidies, that really explains the whole limitations.

Having real-time inflation-protected cash would be the dream for shopping groceries (your bank balance being automatically adjusted for inflation :grin:).

But that would defeat the purpose/correctional effects of inflation, wouldn’t it? I’d guess such a scheme would lead to hyperinflation until the government no longer has an ability to keep its indexation current.

Investing involves risks, the alternative would be a stagnant society, which bears a different but on the long term more disruptive set of risks.

3 Likes

Yes, I fully agree. There’s probably no magical solution to the inflation-issue.

The practical problem I’ve been trying to solve: I do not want to withdraw from my stocks and bonds when both are underperforming. So I’d hold some cash in my FIRE portfolio (great immediate protection for a deflation-shock) for my monthly draw-down/withdrawal. But I still haven’t found the counterpart, immediate inflation-shock protection of your purchasing power.

Some suggest commodities are the way to go for that:

I would say those are two different problems.

For the inflation one: on the short term, normal (non-hyper) inflation has a relatively “tame” effect. If you were planning with a 2% normal inflation and instead end up with, say, 10%, that’s an 8% differential that your cash has to cover over a year. If your risk assessment tells you that you have to cover that risk, you can plan with a 10% buffer on your cash reserve. Additional flexibility comes from cutting expenses, which, for a definite period of time, we usually can do.

Now, once the amount of money you need for living expenses has been majored to account for higher than usual inflation, you can plan your allocation to limit the risk of having to sell assets when they’re depressed.

The first step would be to assess how important for you that risk is by going through the need for risk, ability to take risk and willingness to take risk routine. For example, I have a very high need to take risk, a decent ability and a good willingness to do so. I am willing to take the risk of having to realize losses on my invested assets because the upside would make a real difference in my life, while the downside of loosing it all would not. Your personal assessment seems to be the reverse: if you either have a low willingness, ability or need to take risk, then that is what should drive your investing policy.

I see two ways to mitigate the potential need to sell depressed assets, they come at the sacrifice of some potential gains:

  1. Liability matching: building a ladder of Kassenobligationen or Bonds meant to cover your expenses during the time window that you want. For stocks, I’d take 15 years. You can adapt that using historical data and applying a safety margin to account for whatever period of high inflation and depressed assets price you expect to have to deal with.
    Then, I’d start by the end of that period (let’s say I’m building 6 months levels (each bond lot covers 6 months of expected expenses), I’d buy a lot expiring in 14.5 years, then deduct the interests it would pay during the 6 months prior to assess the size of the lot needed to expire in 14 years, rinse and repeat until I’d reach my minimal duration, which I would finance with a savings/checking account. Whenever a lot expires, if I don’t need it in the immediate future, I’d then assess my expected expenses 15 years from now and buy a new lot expiring in 14.5 years for that amount.
  1. A non-leveraged risk parity portfolio using multiple asset classes (I’d use stocks, bonds, cash and gold. You could add real estate, commodities and/or cryptos if you’d like), designed to get the best possible risk adjusted returns and limiting drawdowns. It would not be perfect, you would still sometimes withdraw assets that have lost value and you would likely have lesser returns than a riskier portfolio but you would likely go through lower drawdowns than other investors and selling your assets shouldn’t hurt as much.

If you get to a point where the situation you are living goes beyond what you had planned for and you can’t get through it by only adapting your lifestyle, that is the kind of situations for which we are building our portfolio and it is a time when it makes sense to deploy (parts of) it, depressed assets or not.

Wealth is relative, if you are loosing wealth but still doing better than 99% of the people around you, chances are your lifestyle will still be comfortable enough to be considered a good life.

2 Likes

Thanks, really appreciate. I’ve had a small epiphany due to your mentioning of the relatively small cost inflation has on cash in the short term.

a) My guess is that any other kind of short-term inflation protection would be more costly than cash, considering the opportunity costs of missed returns from stocks/bonds (e.g. for commodities). What do you think?

b) Instead of a bond ladder, would a global bond fund of different durations (short to long-term, average 7 years, CHF-hedged, e.g. BNDW) be a valid alternative? While not guaranteeing purchasing power, it may provide the same real return as a ladder in the long run, correct?

c) What still confuses me: How do nominal bonds react to inflation (interest rate, bond price)? I guess at some point nominal bonds must increase rates, because else they’d become too unattractive in inflationary times?

d) Any guess on whether the Swiss national bank’s ever going to issue TIPS? As you’ve mentioned above, for retirees (withdrawing from their portfolios) with most of their liabilities being nominal, generally holding more TIPS than nominals in their portfolio might be the better choice?

Again, here we have pensions for this. I mean, SNB has no need to issue products for retail investors. And as it was mentioned in another thread, lots of bonds are held by institutional investors, such as pension funds :rofl:

2 Likes

I agree in general. But you get no pension if you FIRE, or if you decide to cash out your 2nd pillar and invest yourself, living solely off your portfolio in retirement (tiny AHV set aside). That’s my main focus - highest safe withdrawal rate / portfolio lasting as long as possible, hypothetically not depending on any government benefits.

For such a retirement setting, I’ve been asking myself if TIPS would be the better investment than nominal bonds (because inflation poses a greater threat when you do not accumulate anymore and are spending your money mainly for inflation-affected goods).

Maybe double check the theory of what TIPS do, afaik it’s more of building block for hedging than a direct investment (it reacts to inflation expectations changes).

The fact that inflation expectations have been off recently (because the market underestimated it) shouldn’t mean that it’s a good investment (it would be anchoring on recent returns without regard to fundamentals).

The problem, for me, is to combine inflation hedge and a short term horizon. The short term horizon tool is cash, or short dated papers/bills/bonds, which loose value with inflation (but not nearly as much as we’ve seen stocks and long dated bonds do recently with increases in central banks reference rates).

So I’d tend to agree that the short term horizon trumps the concerns in regards to inflation and, as such, for short term needs, be it through inflationary or deflationary times, cash and short term instruments are the way to go.

Bond funds are a different beast altogether. What matters for liability matching is to match the duration of the asset to your liabilities. Bond funds have an “average duration” that can be used to match horizons but they tend to try to keep it constant, while a determined liability comes closer with time, so you would have to use a mix of bond funds with different durations and mix them to match the timeframe of your expected liability/ies as time goes by. I’m not even sure this works since I have yet to see an article/post addressing the issue of using bonds funds as your liability comes due.

Nominal bonds shouldn’t react to inflation directly, but they do react to interest rate raises both actual and expected. This is the best, understandable, information I’ve seen on the topic: Short- and longer-term effects of rising interest rates on a bond fund - Bogleheads.org

The SNB doesn’t issue bonds, the swiss government does (as well as the cantons, the cantonal banks, the municipalities, and corporations).

That being said, I can’t know for sure but I’d say we have a relatively small bond market with low inflation and they haven’t seen the need for it arise. One reason I’ve seen why the US Treasury emits them is to get data on the inflation expectations of the market (the breakeven inflation rate that can be calculated by comparing TIPS and nominal expected returns for the same duration). I guess the SNB either can’t (small market) or doesn’t see the need to get that data.

I would say it is true if you don’t hold them to maturity, in which case, the NAV is greatly affected by expected inflation. When held to maturity, the final face value will reflect the actual inflation adjustment that occurred during its life time.

If they were accessible to us (bonds indexed on the swiss CPI), I would say that they are a good tool to build up the safe part of our retirement savings, knowing that most of our consumption in retirement will be items affected by inflation. They’re not accessible to us, however, so the point is moot (using bonds indexed on other countries inflation is a bet and not an inflation hedge).

2 Likes

But you won’t pay the face value initially right? You pay the market price which reflects the inflation expectations for the time period.

Yes, but when you pay the market price, you know exactly what coupons you’ll receive in real dollars and what amount of principal you’ll get paid at maturity, in real dollars, barring Treasury default.

You don’t get to profit from the conditions effective at the issuance of the bond, but you still know what you get with the least amount of uncertainties possible.

If used to cover liabilities with as much certainty as possibly available, the choice between an inflation protected bond and a nominal one is whether you’d rather have real or nominal dollars.

1 Like

That’s a quite nice explanation actually.

So basically, if I cover my living expenses drawing down on my FIRE portfolio (e.g. 60/40, 600k stocks, 400k bonds, no other income), I’d like to hold some inflation-protected bonds for my inflation-affected liabilities. But surely, I’d never use ALL my bonds for spending purposes, only a very small part of it. So it doesn’t make sense to me to hold more than, say, 1 year of expenses in inflation-protected bonds.

Or does it matter at all :crazy_face:? In the end, I might do better with nominals or TIPS, noone knows. And in the long run, maybe it’ll all reverse to the mean anyway?

I think liability matching takes a different path than a fixed percentage allocation (like a 60/40 one). So, if I understand your framing correctly and you plan to use your bonds for rebalancing purposes, I’d say we enter the “speculation on inflation” territory pointed by nabalzbhf.

In such a scenario, it is the whole portfolio that is used to cover expenses, so the tradeoff would be, as far as I understand it (which isn’t far enough), to choose between slightly higher expected returns (nominals: there’s a premium priced in TIPS for the inflation protection they provide) or some tame protection vs higher than expected inflation.

In all cases, we’re still at the point where no such instrument exists for Switzerland, so unless you plan to have your expenses in the US, the UK or another country offering bonds indexed to inflation, the point remains theoretical and pretty moot.

1 Like