Bonds

[GeoffF100]

My situation is different. I am retired with financial resources far greater than I need. My equity percentage does not matter. I follow another rule of thumb: "if there is not other way to decide go for 60% equities and 40% bonds". The market as a whole is less than 40% equities, so I am heavily overweighting equities. A 60/40 portfolio is possible only if others underweight equities and overweight bonds. There is a danger that I will have to pay too high a price for equities to get others to give them up in favour of bonds. (Actually, I bought most of my equities a long time ago, and I am just sitting on them, but that is another matter.)

This is a nonsense argument. What about other assets such as M2 money? The ratio of M2 money to bonds is about 83:113 so should you hold a respective amount of cash? What about property? Precious metals? Collectables? Crypto?? You can't simply look at the ratio of Global assets and assume that's any indicator whatsoever as to what an investor should do.

You seem to be applying the same faulty passive investing logic that you have used to justify investing in equity trackers. Bit of a man with a hammer?

This has all been explained here before. See, for example:

https://www.investopedia.com/terms/m/ma ... tfolio.asp

It is not practicable to buy the market portfolio. Nonetheless, the market portfolio is fundamental to understanding market pricing, and how to build a diversified portfolio. The concept was invented by Nobel Prize winners, not me.

[Bubblesofearth]

This has all been explained here before. See, for example:

https://www.investopedia.com/terms/m/ma ... tfolio.asp

It is not practicable to buy the market portfolio. Nonetheless, the market portfolio is fundamental to understanding market pricing, and how to build a diversified portfolio. The concept was invented by Nobel Prize winners, not me.

So do you believe that an ideal portfolio for a long term individual investor is to hold about 40% of their money in cash?

BoE

[JohnW]

All the hand wringing about an unprecedented bond rout in 2022 could have been avoided with duration matching our bond investments, which is how you minimise duration risk with bonds;

So what duration would have avoided the 2022 rout then?

Proceeding with trepidation.....I think UK central bank interest rate rose 5% over that particular three year period, and 10 year gilt yield rose 4.5%, but most of them happening over two years and a lot during just 2023.
Vanguard has a couple of short duration funds that lost about 6% and 16% then (peak to trough). I don't see their duration but the majority of the holdings are less than 3 years maturity.
If that duration is too long for an investor can you hold some cash, call it bonds, which has a zero duration, in a suitable proportion with your actual bonds to get a duration that suits you?

[Gpop321]

It's literally "own your age in bonds, the rest in equities."

...as a gentle introductory pointer to the role of bonds and equities tho, not as a hard and fast rule.

Definitely agree with your point that 100 years stats aren't bullet proof, but it makes more sense than trusting a smaller sample period, surely?

That is 100 years stats for the two most successful markets, during a strong period of economic growth. The next 100 years for the global market will not necessarily mirror that. Apart form the current high valuations, we have climate change, a war brewing in Europe and perhaps in the far east. We can only say that we do not know. The optimists won over the last 100 years, the pessimists may win over the next 100 years. How much money do you need to make? What is the lowest risk way of making it?

I both totally agree that the next 100 years could throw all types of hell at us. Equally tho, the last 100 saw WW2, the Cuban missile crisis, the Cold War, 9/11 etc etc etc, yet the net result was a continuous growth of the world market. The only certainty for the next 100 years is, I think, that there will be multiple catastrophes along the way. Our only inidications of how the markets might weather that come from how they weathered the rather hellish previous 100 years. If we don't use that as some sort of guide then we are in effect guessing/hoping, are we not?

On the final sentence: I've got a target amount in mind, and a timespan within which I want to achieve that. Lowest risk of doing that tallies with my risk tolerance, which is just beyond the midpoint (on the risky half).

[dealtn]

So what duration would have avoided the 2022 rout then?

Proceeding with trepidation.....I think UK central bank interest rate rose 5% over that particular three year period, and 10 year gilt yield rose 4.5%, but most of them happening over two years and a lot during just 2023.
Vanguard has a couple of short duration funds that lost about 6% and 16% then (peak to trough). I don't see their duration but the majority of the holdings are less than 3 years maturity.
If that duration is too long for an investor can you hold some cash, call it bonds, which has a zero duration, in a suitable proportion with your actual bonds to get a duration that suits you?

So in effect not duration matching at all then. Simply investing at the (extreme) short end of the curve.

[Alaric]

So in effect not duration matching at all then. Simply investing at the (extreme) short end of the curve.

Duration matching would be to project each future year's required income, and invest in bonds such that in any future year the maturing bond in that year plus the coupons on non maturing bonds would be sufficient to meet the income requirement. With that in place, interest rate wobbles have no particular effect other than the change the return on any cash that needs to be reinvested or give a gain or loss if any bond has to be sold before maturity..

[dealtn]

So in effect not duration matching at all then. Simply investing at the (extreme) short end of the curve.

Duration matching would be to project each future year's required income, and invest in bonds such that in any future year the maturing bond in that year plus the coupons on non maturing bonds would be sufficient to meet the income requirement. With that in place, interest rate wobbles have no particular effect other than the change the return on any cash that needs to be reinvested or give a gain or loss if any bond has to be sold before maturity..

No. You are still exposed to change in duration risk. And also event risk.

You (or a partner) get a medical diagnosis, you get divorced, a change in tax treatment, change in inflation definition (for linkers), plus multiple other examples.

If you have invested for a 30 year duration, and something happens when it is now appropriate to have 10 (or vice versa), you are exposed to curve risk. In 2022/23 investing in 30 years and losing perhaps 50% and switching would have been painful.

[GeoffF100]

This has all been explained here before. See, for example:

https://www.investopedia.com/terms/m/ma ... tfolio.asp

It is not practicable to buy the market portfolio. Nonetheless, the market portfolio is fundamental to understanding market pricing, and how to build a diversified portfolio. The concept was invented by Nobel Prize winners, not me.

So do you believe that an ideal portfolio for a long term individual investor is to hold about 40% of their money in cash?

Why would I believe that? There are some who advocate investing the market portfolio (as far as is practicable) rather than just studying it to gain understanding, e.g.:

https://portfoliocharts.com/portfolios/ ... allocation

[GeoffF100]

Our only inidications of how the markets might weather that come from how they weathered the rather hellish previous 100 years. If we don't use that as some sort of guide then we are in effect guessing/hoping, are we not?

Why pick 100 years? Why not 200? Why pick one market and not another? We are guessing/hoping whatever we do.

[JohnW]

Proceeding with trepidation.....I think UK central bank interest rate rose 5% over that particular three year period, and 10 year gilt yield rose 4.5%, but most of them happening over two years and a lot during just 2023.
Vanguard has a couple of short duration funds that lost about 6% and 16% then (peak to trough). I don't see their duration but the majority of the holdings are less than 3 years maturity.
If that duration is too long for an investor can you hold some cash, call it bonds, which has a zero duration, in a suitable proportion with your actual bonds to get a duration that suits you?

So in effect not duration matching at all then. Simply investing at the (extreme) short end of the curve.

No, I don't think that follows from my observations. You've misread my 'hand wringing can be avoided' as 'rout can been avoided'; it's easy to do, and very tempting sometimes, but does add some piquancy. It is as you say investing with extreme short duration which will avoid the hand wringing (since the value fall will be small) if you've duration matched the bonds to near term spending. With much longer duration bonds we will and did see the rout, but if your spending duration is far into the future who gives a toss how far prices fall now, because they'll recover faster than otherwise now that bond yields are that much higher than they were? Rather than wringing, hand rubbing with excited expectation.

[Bubblesofearth]

Why would I believe that? There are some who advocate investing the market portfolio (as far as is practicable) rather than just studying it to gain understanding, e.g.:

https://portfoliocharts.com/portfolios/ ... allocation

I thought your argument was that investing in the market portfolio was optimal? If not then where do you draw the line? No cash despite the large large weighting of that asset class? And what about by far the biggest Global asset - property? Last I looked the Global property market was almost 3X the size of the Global bond market.

Understanding the market portfolio is fine but gives no a priori reason to invest in it. Same for equities when considered in isolation - understanding how passive trackers work doesn't mean you should invest in them.

BoE

[GoSeigen]

So in effect not duration matching at all then. Simply investing at the (extreme) short end of the curve.

No, I don't think that follows from my observations. You've misread my 'hand wringing can be avoided' as 'rout can been avoided'; it's easy to do, and very tempting sometimes, but does add some piquancy. It is as you say investing with extreme short duration which will avoid the hand wringing (since the value fall will be small) if you've duration matched the bonds to near term spending. With much longer duration bonds we will and did see the rout, but if your spending duration is far into the future who gives a toss how far prices fall now, because they'll recover faster than otherwise now that bond yields are that much higher than they were? Rather than wringing, hand rubbing with excited expectation.

Ahem, in 2006/7 there was no "hand rubbing with excited expectation" and yields were higher than they are now. I know because I was practically a lone gilt investor at that time and everyone told me I was nuts because "the only way for yields was up" and they were "definitely going to 10%".

So tell me, what has changed so much that bond yields now cannot go to 10% and that we are hand rubbing in excited expectation of yield falls?


I agree with you about duration matching by the way. Buying more or less duration than required is a often good way to get disappointing returns.


GS

[GeoffF100]

Why would I believe that? There are some who advocate investing the market portfolio (as far as is practicable) rather than just studying it to gain understanding, e.g.:

https://portfoliocharts.com/portfolios/ ... allocation

I thought your argument was that investing in the market portfolio was optimal? If not then where do you draw the line? No cash despite the large large weighting of that asset class? And what about by far the biggest Global asset - property? Last I looked the Global property market was almost 3X the size of the Global bond market.

Understanding the market portfolio is fine but gives no a priori reason to invest in it. Same for equities when considered in isolation - understanding how passive trackers work doesn't mean you should invest in them.

The concept of the market portfolio stems from CAPM. Developed equity markets fit the assumptions of CAPM well. With the assumptions of that model, the market portfolio is optimal for equities. Other asset classes fit the model less well, or not at all. Nonetheless, the concept market portfolio can still be useful for reasoning about the asset classes that are a better fit.

My point was that to make my 60/40 portfolio possible others have to underweight equities and overweight bonds. They would not do that if they thought that equities were the superior investment. I have not said that the market portfolio bond percentage is optimal for all investors, because bonds do not match the assumptions of CAPM particularly well. In particular, investors do not have have the same objectives. They have different tolerances for risk.

[dealtn]

With much longer duration bonds we will and did see the rout, but if your spending duration is far into the future who gives a toss how far prices fall now,

Most people I imagine. Even those that don'y literally "mark to market" their portfolio valuations like institutions (and professionals) are still likely to be a mixture of disappointed and horrified to have lost money, and being (ir)rational human beings considering cutting losses. Add in the likely reality that higher than anticipated yields (and lower than expected bond prices) is running parallel with inflation and elevated goods and service prices, which are unlikely to reverse, that huge drop in capital, even if/when it recovers in nominal terms, won't be worth close to what it was expected to be in real terms.

Its probable you will be disappointed on two fronts. You have to bear the reality of a drop in capital value, and the uncertain timeframe on when that might/will recover, plus the likelihood that when it does inflation has eaten away at your plans such that your planned spending match of duration is now only partially met.

[GeoffF100]

Why would I believe that? There are some who advocate investing the market portfolio (as far as is practicable) rather than just studying it to gain understanding, e.g.:

https://portfoliocharts.com/portfolios/ ... allocation

I thought your argument was that investing in the market portfolio was optimal? If not then where do you draw the line? No cash despite the large large weighting of that asset class? And what about by far the biggest Global asset - property? Last I looked the Global property market was almost 3X the size of the Global bond market.

Understanding the market portfolio is fine but gives no a priori reason to invest in it. Same for equities when considered in isolation - understanding how passive trackers work doesn't mean you should invest in them.

You can draw practical conclusions from CAPM only where its assumptions apply sufficiently well to justify those conclusions. If you want answers to your questions, you will need to study CAPM and the other associated financial theory.

[JohnW]

Most people I imagine. Even those that don'y literally "mark to market" their portfolio valuations like institutions (and professionals) are still likely to be a mixture of disappointed and horrified to have lost money, and being (ir)rational human beings considering cutting losses. Add in the likely reality that higher than anticipated yields (and lower than expected bond prices) is running parallel with inflation and elevated goods and service prices, which are unlikely to reverse, that huge drop in capital, even if/when it recovers in nominal terms, won't be worth close to what it was expected to be in real terms.

Its probable you will be disappointed on two fronts. You have to bear the reality of a drop in capital value, and the uncertain timeframe on when that might/will recover, plus the likelihood that when it does inflation has eaten away at your plans such that your planned spending match of duration is now only partially met.

You're likely right, most people. But we might help folk avoid behavioural mistakes by recommending they view their portfolio as a whole (stocks didn't do too terribly when bonds went on a rout in 2023), and by taking the long view for a long term bond holding (when prices fall you're getting a bargain if you're buying more bonds).
As to the hazards of inflation during a long term bond recovery, well it takes courage to hold only nominal bonds for the long term when linkers will do the job and give better inflation protection. Nominal bonds for the short term during which inflation can't do too much wrecking, but not too many nominals for the long term I'd suggest.

[Bubblesofearth]

You can draw practical conclusions from CAPM only where its assumptions apply sufficiently well to justify those conclusions. If you want answers to your questions, you will need to study CAPM and the other associated financial theory.

I've studied it and found what I believe are significant flaws which I've outlined previously with examples. However, I recognise there is no way you will accept anything I say that goes contrary to the wisdom of the Nobel winners.

BoE

[Alaric]

[ have not said that the market portfolio bond percentage is optimal for all investors, because bonds do not match the assumptions of CAPM particularly well. In particular, investors do not have have the same objectives. They have different tolerances for risk.

Not just risk but some investors in bonds such as Insurance Companies and Pension Funds will want to match by duration, thus constraining thir ability to sell or switch.

[JohnW]

Why would I believe that? There are some who advocate investing the market portfolio (as far as is practicable) rather than just studying it to gain understanding, e.g.:

https://portfoliocharts.com/portfolios/ ... allocation

I thought your argument was that investing in the market portfolio was optimal? If not then where do you draw the line? No cash despite the large large weighting of that asset class? And what about by far the biggest Global asset - property? Last I looked the Global property market was almost 3X the size of the Global bond market.

Understanding the market portfolio is fine but gives no a priori reason to invest in it. Same for equities when considered in isolation - understanding how passive trackers work doesn't mean you should invest in them.

BoE

One might deviate from it for a host of reasons: it might be difficult/impossible/expensive to invest in some of it, like some property; one might prefer more returns even if the risk is disproportionately higher, like 'all equities'; you might be excluded from some segments like private equity; one might not be able to accept the risk, but only tolerate the risk of short term bonds.
So it becomes, I think, a non-sequitur to suggest that because we know how good the market portfolio is but don't invest in it, that there is no good reason to invest in cap weighted global equities just because we know the returns match the market returns.

[GoSeigen]

Most people I imagine. Even those that don'y literally "mark to market" their portfolio valuations like institutions (and professionals) are still likely to be a mixture of disappointed and horrified to have lost money, and being (ir)rational human beings considering cutting losses. Add in the likely reality that higher than anticipated yields (and lower than expected bond prices) is running parallel with inflation and elevated goods and service prices, which are unlikely to reverse, that huge drop in capital, even if/when it recovers in nominal terms, won't be worth close to what it was expected to be in real terms.

Its probable you will be disappointed on two fronts. You have to bear the reality of a drop in capital value, and the uncertain timeframe on when that might/will recover, plus the likelihood that when it does inflation has eaten away at your plans such that your planned spending match of duration is now only partially met.

"Losing money" with bonds is a funny concept -- not to minimise the recent crash and the pain of anyone who might foolishly have been speculating on gilt prices (more fool them).

IMV JohnW is right about the duration matching. Anyone who duration matched properly is not likely to have "lost" much. Think about people investing 2ish years ago (pre-crash):

Let's start with people who wanted their money in two years (today) for example. It they bought a two-year gilt they did so at a yield of (say) zero percent and at the end of the two years their return was zero. Did they lose anything? They were stupid to buy gilts with such a poor yield in the first place, but they knew what the return would be and they got that return. Okay, they missed out on the chance to invest their capital at 4% for a few months but that is no great loss -- their money was tied up anyway by that initial decision to invest - it is not the price falls that tied them in.

Now what about someone with a longer horizon, say ten years. If they bought a ten-year (duration) gilt, similar argument to the above. The return is around 0% but that is what they signed up for.

What if the investment horizon was ten years but they bought a 30-year gilt? Well, presumably then a third of their portfolio was invested in the gilt but the remainder in cash resulting in a portfolio duration of ten years (or an equivalent arrangement). Now of course if yield remain unchanged the thirty-year bond shows a significant loss at the end of the ten years. BUT the 66% cash allocation which was expected to earn 0% over the ten years now earns 4%+ (or can immediately be invested in 8-year gilts earning 4%). These 4% compounding interest payments over the remaining eight years will largely recoup the loss on the 33% thirty-year gilt component. Again the return over the ten years is close to zero but that is what was expected at the outset.

So to me duration matching is sound and maximises the probability of meeting the investment objective. Note that if the duration did not match then the outcome would be more divergent e.g. the two-year investor putting 100% into 30-year gilts (?!) would have a huge loss.



This is Bond 101 so apologies if it's obvious to anyone, but I'm struggling to see how dealtn is taking this on board.


GS