Third each bond/stock/commodity - conservative asset allocation

[1nvest]

Asset allocation :

Calendar year rebalanced

1825 to 1931 25% UK stock, 25% US stock, 50% cash. When the Pound was on a metallic standard (silver/gold), directly convertible at a fixed rate, it was reasonable to hold cash - invested in a bank/deposit account earning interest (UK T-Bills).

In late 1931 the UK broke off the gold standard (too much money was being converted for gold and the gold being removed from the country, a similar reason why the US broke off the gold standard in 1971). Asset allocation modification to 17% UK stock, 17% US stock, 33% cash (T-Bills), 33% silver (as US was still on the gold standard).

1972 onwards, after the US had broken off the gold standard, the same as the above, but with gold instead of silver.

Exceptions : All silver held during periods of World Wars (uncertainties of outcomes).

So since 1972 a asset allocation of a third each in gold, UK cash/T-Bills, and a 50/50 US/UK stock blend.

Performance :

After inflation total return (accumulation)






Notes : A more recent (since 1972) form of half weighting into a 33% US stock (US$ primary reserve currency), gold (global currency) blend, and half weighting into 33% UK stock and UK T-Bills (cash) (£). Neutral currency stance (half £, half foreign). Whilst holding a third each in stocks, bonds (T-Bills/cash) and commodity (gold).

Base data (note that 1871 to 1895 UK dividend yield data is n/a, so in the above charts I assumed a constant 4% UK stock dividend yield for those years).

Caveats :

Data is sourced from a wide range and the validity is unknown.

Gross total return values - excludes costs/taxes.

Suggestions :

Apply a low/safe SWR of 2% to provide part of income in a regular inflation adjusted manner, top slice additional real gains as and when apparent into a separate cheque account that is drawn from in a discretionary manner to supplement SWR income. Often fast/large gains will occur and top slicing some/all of those gains is appropriate.

Look to rebalance at/around fiscal year end (5th April) as the flexibility that provides in opting to rebalance in the old, new or a combination of both - whichever might be the more tax efficient.

Hold some of gold in physical, some in gold fund (lower trading costs)

Definitions :

SWR = the initial percentage of the total start date portfolio value that is drawn as income, and where that £ amount is uplifted by inflation as the amount drawn in subsequent years.

Notes :

Historically across all 50 year periods in the above data (calendar year granularity), income production would have been of the order 4% minimum, 5% average, whilst still preserving the inflation adjusted start date portfolio value. i.e. a common reward as per the widely anticipated 4% SWR 'rule', whilst bad years were relatively few and lost relatively little (a relatively comfortable year on year portfolio).

Summary :

Primarily a currency neutral, half £ half foreign asset allocation, diversified equally across stocks, bonds and commodities (gold is a form of global currency as well as being a form of commodity index holding). Where historic data/assets are revised towards what might more reasonably have been held in the past in reflection of circumstances evident at the time.

[Bubblesofearth]

Summary :

Primarily a currency neutral, half £ half foreign asset allocation, diversified equally across stocks, bonds and commodities (gold is a form of global currency as well as being a form of commodity index holding). Where historic data/assets are revised towards what might more reasonably have been held in the past in reflection of circumstances evident at the time.

If by bonds you mean US T-bills or UK gilts then I see no point in private investors holding these at the moment.

If you are given the choice of two assets with the same approximate expected return but one has volatility (risk) associated with it and the other does not then why hold the volatile one? Cash makes more sense than bills/gilts just now.

BoE

[tikunetih]

If you are given the choice of two assets with the same approximate expected return but one has volatility (risk) associated with it and the other does not then why hold the volatile one?

To harvest rebalancing premium would be a reason.

[Bubblesofearth]

To harvest rebalancing premium would be a reason.

Only if they are negatively correlated and the evidence for that is weak at best.




BoE

[1nvest]

Use the data in the first post to compare UK share price only gains with cash (UK T-Bills/gilts), broadly the two align (same/similar reward), but in a volatile manner. Rebalancing between high and low volatility assets that have similar broad rewards tends to yield a higher reward than just holding one or the other alone. But yes the difference between cash and UK stock price only 'rebalance benefit' was small, of the order 0.4% - likely due to relatively high correlation. That also excludes stocks throwing off a average 5.4% dividend benefit on top, such that the 'rebalance benefit' was considerably less than dividends.

Revise the calculation to add dividends to cash each year (so cash increases by the interest it earns + stock dividend cash the stock holdings generate) and again yearly rebalance back to 50/50 and the overall difference between stock total return (dividends reinvested/accumulation) and 50/50 stock/cash was 2% annualised (6.2% annualised real for all-stock, 4.2% annualised real for 50-50).

That's with the safest form of cash, short dated gilts (year or less), fully depositor protected no matter how much is 'deposited' (unlike Banks where only the first £85,000 at present is protected). Likely you could have improved upon that, perhaps even a couple of percent better on average - which would have closed the gap down to perhaps 1% difference between all-stock and 50/50.

The rolling 5 year returns indicate greater volatility in stocks, often negative even in nominal terms; With cash the rolling 5 year average tends to be smoother. 50/50 will tend to smooth down the volatility compared to all-stock.

Mix in another asset - gold/currency and again that will see swings - that can be traded (such as via yearly rebalancing). Cisco for instance list in both the US and Mexico, whilst the share value/price is the same for both of those, you also have the currency volatility also mixed in, stock total return +/- currency change.

For simplicity, resorting to US data and for the longest amount of data available (from 1972) to recent, a third each in US stock/gold/Gilts (US treasury) compared to 100% all-stock, to recent, has seen a 0.5%/year difference in total returns. https://tinyurl.com/y8q2u52n Within that, as all stock is more volatile, it will have zigzagged around more, some periods within that seeing much better results, other periods worse results. As a bad case example, starting in April 2000 to February 2009 https://tinyurl.com/y9da6zbt and a investor in all stock who was drawing 4%/year SWR saw their portfolio value in real (after inflation) terms decline down to around 25% of its start date value, £10,000 start date value declined to £2530 in inflation adjusted terms (in that link you have to hover your mouse/pointer over the Portfolio Growth chart lines after having set the chart to being 'Inflation Adjusted' to see those values). In contrast the same 4% applied to the stock/gold/cash blend saw a decline down to £9426 (near-as - no loss).

If after around a decade you'd been drawing/spending 4% inflation adjusted 'income' out of total returns from a all stock portfolio, and the value of that portfolio had declined 75% in inflation adjusted terms, the prospect of the longer term historical average reward suggesting a 0.5%/year more annualised return for all-stock compared to a third each stock/bond/gold would be pretty much irrelevant to you. The realities being that a large chunk of decades of saving/investing could have been lost, and down to levels where forward time prospects looked bleak. Many would capitulate at such time, looking to preserve what little remained. Only if you had nerves of steel might you have persisted with that asset allocation.

Sequence of returns matters, and is one of the primary factors of how well, or not, your actual outcome is. If bad things happen to the portfolio in earlier retirement (drawdown) years more often that has much greater impact compared to if bad things happened in later years. When portfolio value is smoothed down via diversification across assets/currencies then that risk can be substantially reduced, but at perhaps a broader average 0.5%/year type insurance cost.

Yes I've only looked at the bad side, another example could equally be selected where the all-stock gains were far far better than the stock/gold/bond blend, but even in that case the stock/bond/gold blend rewards will likely still have been reasonable/good, just not as great as all-stock. Regret cost (not having been all in stock to instead have held stock/bond/gold across a period when stocks were outstanding is actually a low-risk factor if the portfolio has still met its objectives of providing a sufficient retirement income), compared to the risk of a large decline and capitulation type outcome as I outlined earlier.

Cash makes more sense than bills/gilts just now

Depends upon liquidity and scale factors. Yes if I can earn 2%/year more by buying High Street Bank fixed income bonds and the amounts are within the depositor protection limits (£85,000 with any one bank/group), then that seems the clearly better choice, assuming I am also rebalancing as/when each bond matures (and the proceed amounts compare to how much might be being indicated to be deployed into stock/whatever at that time). But at the cost of liquidity - for instance if stocks dropped a lot such that it seemed appropriate to liquidate some 'cash' in order to buy more shares. Fixed income bonds have fixed terms, cannot be liquidated quickly and likely with a penalty if closed early. And if your cash exceeds depositor protection levels then Gilts also have the benefit of being fully depositor protected, no matter how much your 'deposit'.

IMO a key part of investing is focusing upon the 'cash' assets - bonds/high street fixed income ...etc. At times of relatively high yields you should be looking at locking into holding some longer term bonds, at other times shortening down the duration/maturity so that if/when interest rates do spike back up you roll into those higher yields without having lost (or lost relatively little) capital value. Increasing reward moderately whilst increasing risk marginally from 'cash' (bonds) can if managed well, close down the difference between a more diversified portfolio compared to all-stock, maybe even totally, maybe even more.

[1nvest]

A 'Suggestion' I missed in the original post is that if you're investing a lump sum then do so over two years, three time points (day zero, after a year, at the end of the second year (or perhaps 1/24th added monthly ). Historically that uplifted (improved upon) the worst case outcome. Avoided having gone all-in at the worst possible time.

[1nvest]

Many have been suggesting to avoid mid/longer dated gilts for years now - pretty much since the 2009 financial crisis and transition over to low yields. However long dated (20+ year) gilts have presented opportunities for 25% type capital gains at times, including during the last couple of years. For instance this is a recent price chart for a 20 year gilt ...

Back in the 1970's (? 1980's), you could buy into long dated double digit longer dated yields. Looking back and it looked like a no-brainer to do so, but many didn't - due to fear of even higher yields/lower prices to come. One way to better ensure that you did buy at least some is to use a barbell of short and long dated gilts instead of a 10 year bullet, as that way your yearly rebalance will at times at least have had you buy some of the longer dated. Yes they'll take a hit over some periods and be a drag factor, but over the mid to longer term that can yield dividends. It's the same for gold, over some periods the price will trend down, but rebalancing will tend to have you buying more ounces of gold, and at a later date having expanded the ounces of gold being held will pay back in the manner of deploying/selling some gold to add more stock shares to the portfolio. Those cycles tend to be long, potentially spanning multiple decades. Across the 1980's/1990's for instance whilst the nominal price of gold declined around 33%, 50/50 with stock and yearly rebalanced saw you accumulate 6 to 8 times more ounces of gold, and the benefit of such accumulation, whilst a drag factor across those 1980's/90's years, later provided dividends in the way of rebalancing the other way around, selling ounces of gold to buy more stock shares at relatively low price levels, and/or more Pounds after the £ declined.

There are three main sources of rewards, price appreciation, income (dividends/interest) and volatility trading. Trading volatility can be as simple as via yearly rebalancing - largely automatic. Which of those is the more rewarding? Well conceptually they should all compare equally. Growth investors look to price appreciation, HYP investors look towards income, Options traders often look towards volatility capture. None of those consistently is the better, as if that were the case then all investors would focus in on that single best case choice. Diversifying across exposure to all three is the more balanced portfolio choice.

Similarly stock/bond/commodities will each have their time both with and against the tide, diversification again helps. As does holding some foreign currency to account for the Pound rising/falling over time.

50/50 Pound/foreign (third each stock, UK bond, gold, with stock 50/50 US and UK ... i.e. 16% in £ (UK) stock, 33% UK bonds, 50% combined; Alongside 16% primary reserve currency (US$) stock, and 33% gold (global currency).

Third each in stock, bonds (cash) and commodity (gold).

Along with yearly rebalancing to 'trade' the fluctuations/volatility.

... simple to set up and manage. And to (optionally) potentially enhance rewards - focus on the 'cash' (bond) element. If you can enhance your 'cash' rewards to be greater than the returns provided by short term gilt (T-Bills) in a low risk manner, then that enhances overall portfolio rewards. Likely that will be a lot easier to achieve than attempting to enhance stock rewards.

[Bubblesofearth]

Cash makes more sense than bills/gilts just now

Depends upon liquidity and scale factors. Yes if I can earn 2%/year more by buying High Street Bank fixed income bonds and the amounts are within the depositor protection limits (£85,000 with any one bank/group), then that seems the clearly better choice, assuming I am also rebalancing as/when each bond matures (and the proceed amounts compare to how much might be being indicated to be deployed into stock/whatever at that time). But at the cost of liquidity - for instance if stocks dropped a lot such that it seemed appropriate to liquidate some 'cash' in order to buy more shares. Fixed income bonds have fixed terms, cannot be liquidated quickly and likely with a penalty if closed early. And if your cash exceeds depositor protection levels then Gilts also have the benefit of being fully depositor protected, no matter how much your 'deposit'.

Why do you need to earn 2% more with cash? UK short gilt yields are currently negative and longs yielding only about 0.5%. You don't need to have cash on fixed interest deposit to be as good as gilts! Factor in the cost of buying and selling gilts and they are utterly without merit for private investors just now.

OK, if you are mega-wealthy there may be some issues with cash but most of aren't that wealthy. You can spread the money between banks/building societies to keep the protection.

The fact that gilts have been in a bull market is irrelevant, it's the future that matters.

BoE

[1nvest]

OK, if you are mega-wealthy there may be some issues with cash but most of aren't that wealthy. You can spread the money between banks/building societies to keep the protection.

Personally I do rate tart around the banks/BS and run a 5 year ladder of such holdings, spread across providers so as to remain within £85K limit (x 2 of us, so £170K). And time portfolio rebalance reviews to coincide with when one rung matures each year. More recently I've been shortening that down to a 3 year ladder as the differences between 3 and 5 years were marginal.

[1nvest]

Just for the record, fiscal 2019/2020

US stock (S&P500) -18% in US$ terms (total return), -6.5% in £ terms (Pound declined from 1.3177 down to 1.2395)
FT All Share -24% (total return)
Gold +34.5% (£978.66 3pm price fix 4th April 2019 to 1315.97 3pm price fix 3rd April 2020)

Assuming 0.5% for 'cash' then 17% US stock, 17% UK stock, 33% gold, 33% cash = +6.3% portfolio return for the fiscal year 2019/2020

50/50 £/foreign trades such volatility. If you're 50/50 £/foreign and the £ halves, you rebalance and then the £ doubles back up again, and you rebalance again, then you're up overall despite the £ being back to where it was before.

Distinctly 2019/20 fiscal year was a Pound down period, and hence a gold up year (along with US stock declines/losses being diluted by stock + currency combined outcomes).

I'd guess, but not bet on, 2020/2021 perhaps seeing stocks up, gold down, £ up. Such that UK stocks might reward more than US stocks after US stock gains are diluted by the £ rising; And where gold was also down. But where gold losses were offset and more by the combined stock gains. But that is purely a guess, often you see a series of gold being the years best performing asset for a run of years, and then stocks being the years best performing asset for a run of years. Just take what's given. Often the years best performing asset is up +20% or more, whilst the other assets collectively are around break-even.

I conjecture that the reason taking a low SWR is beneficial is that it takes some of those fast/larger gains off the table as they occur, which is better than not (letting it ride). Such that overall you tend to see the combined sum of SWR + surplus real (after inflation) gain being larger than just real gain alone (with no SWR being drawn). If you supplement that by taking some/all of surplus real gains also off the table as/when apparent, then that sidelined cash will help better sustain your income/spending. Often you'll see a large chunk of top-sliced real gains loading up your cheque/cash (spending) account, and then a series of years with just the SWR being provided (when you draw down that cheque account), before another largish chunk of cash is thrown off again (portfolio value is back into making new higher highs in real terms after discounting the SWR).

2% is a good choice of SWR IMO, perhaps enough to cover basic living expenses, and leave it to the portfolio to throw off additional cash/spending out of real gains of a further 2% to 3% (so 4% to 5% in total). Which is a form of also diversifying income provision.

If you draw for spending all of real gains, then mid to longer term the portfolio will likely just preserve its inflation adjusted start date value. If you draw only some, let some ride, then the portfolio value will tend to rise in inflation adjusted terms over time - potentially leaving heirs even more. The primary risk/concern is the rate of additional real gain top slicing/cash throw-off above SWR as that is irregular and discretionary. The more stable/consistent the portfolio is over time, the more regular that real gain cash throw-off tends to be, and the portfolio described has a history of portfolio real gain value/growth being relatively stable/consistent.