Basket of Eight: 2022 review

[Luniversal]

The Basket of Eight (B8) was devised in late 2010 for the ignorant and apathetic investor who needs to pay bills as they fall due. The portfolio seeks a flow of income from UK equity-based investment trusts whose purchasing power should at least be constant over time. Operations should as far as feasible be 'fire and forget'. Only one decision has been required of holders since the launch-- see below*.

B8 members were to be bought in equal amounts: City of London (CTY), Dunedin Income & Growth (DIG), Edinburgh (EDIN), Invesco Income Growth (IVI), Merchants Trust (MRCH), Murray Income (MUT), Schroder Income Growth (SCF) and Temple Bar (TMPL).

*IVI was absorbed by Invesco Select Trust and has been replaced by Aberdeen Equity Income (AEI). Data from the retroactively calculated B8 launch date has been adjusted accordingly. Actual results suppose that AEI was bought to replace IVI on Apr. 21, with dealing costs provided for. See the 2020 review update [1]:
viewtopic.php?f=54&t=25362&p=405863#p405863

(Aberdeen Standard Equity Income was renamed on Apr. 6.)

Latest results are for financial years closing between Aug. 2021 and Jun. 2022. Results are aggregated to a common Jun. year end, since this best fits accounting dates. Trends since the B8's backtested launch on Nov. 10, 2000 (also when pyad's 'HYP 1' began) are reviewed.


INCOME
The B8 was clobbered getting off the canvas in 2021-22. The covid panic had devastated dividends from spring 2020. Just as restitution was under way the Ukraine war broke out in Feb. 2022. Inflation has rocketed, cutting payouts' real worth by more than Chinese bugs did. The two-year whammy damaged UK equity income more than anything in living memory.

No trust encountered self-made troubles to compare with Edinburgh's and Temple Bar's previously, but none of the eight managed to increase its dividends by anywhere near Retail Prices Index inflation of 12.3% in the year to Jun. That was four times average inflation in the previous four decades.

When setting divi rates directors barely acknowledged such erosion. They were content to offer small nominal increases, gilding their dubious glory as 'dividend heroes' according to the trade body's money-illusionary criterion. Yet the B8's income performance was disappointing, not disastrous. Regular dividends per share fell by 7.5% (2020-21: down 1.8%) after retail in real terms. Taking the two years together, this is less dire than with many directly held portfolios of shares, e.g. HYPs.

In purchasing power, the B8's income stayed almost one-fifth higher than in its first full year, 2001-02, though it had been one-third higher in 2019-20. The trusts' revenue reserves have done what they are meant to do. They were milked less than in the covid year because boards reined in rises-- although this could be because more turbulence is feared.

Last week's preliminary report from Murray Income, which had sounded optimistic a year ago, was franker than the rest about seeing a medium-term outlook subdued for dividends labouring in the newly inflationary climate. OTOH last week also saw a fairly blithe prognosis from City of London, the archetypal all-weathers navigator for more than 30 years in manager Job Curtis's hands.

The B8 has taken a real decline in income of 1.1% pa over the decade since the Global Financial Crisis played out, during which it averaged a yield of 4.2%: about half a percentage point over the FT All-Share Index (FTAS). The two stats go together: the 'juicier' a basket is at first, relative to the market, the less likely it is to grow that income spiritedly. As a substitute for a fixed interest security or deposit account, it has been satisfactory. Further, real earnings per share over the same decade showed a fractonal 0,1% increase, so reserves have strengthened overall; the trusts have not capitulated to investors' hunger for income in the 21st century by over-disbursing it.

The portfolio's yield on Fri. night-- based on historic or officially forecast payouts and £75,000 gross invested-- was 5.0% (Jun. 2021: 4.4%, Jun. 2020: 5.6%). That hovers between gloom when fear of covid was raging and hope before Putin struck. The B8 yields a point more than the All-Share, pretty much par for its lifetime. It remains sufficiently juicy for starters.

Prudent dividend rises protected revenue reserves. Distributed income was almost covered 0.98 times (2020-21: 0.72x, 2019-20: 0.99x) against a lifetime average of 1x. Reserves were mulcted previously in three of 18 full years, so the current drain is the most prolonged; it has left reserves at a B8 nadir, 70% of current payouts. However three of eight members covered their payouts last year and the rest are getting there. We have heard no talk of realised capital gains being used to top up portfolio receipts either.

One must not be too complacent. The average revenue reserve shrank from 9 to 8 months of current payout in the composite year: it peaked soon before the Global Financial Crisis's fallout hit 2009-10, and has been static or shrinking ever since. A year's worth of the latest payout in hand is customarily regarded by trusts with a progressive policy as adequate. Dunedin and Schroder Income have 15 months, and Edinburgh 14 after cutting. The rest are below the 12-month Plimsoll Line.

The basket's Ongoing Charges Ratio defied a drop in net assets per share: at 0.49% it is at its lowest, down from 0.51%, or 0.63% throughout the B8's span. The number of shares in issue swelled by 4% last year, enabling costs to be diffused across a wider base. Retail appetite for income-paying funds continues to grow: there are almost one-third more B8 shares in issue than in 2011-12. Demographics of retiring Boomers ought to keep them in demand.

Revenue expenses, which come off distributable income, inched up from 4.9% of the payout to 5.1%, well below the lifetime 6.5%. This ratio has lain in the 5-7% range since c. 2008, though tending to reduce as expenses are loaded more on to the capital account.

Dividends per share since the B8's first full year compounded at 4.2% pa against inflation of 3.1% [2]. In the post-GFC decade trusts imposed real cuts, year on year, on 34 of a possible 80 occasions, averaging 4.2% real. Temple Bar's butchery skews the average, though it began to redeem itself with a 7.4% hike in money terms last year: best in breed, if subinflationary.

In purchasing power terms dividends have declined in nine of 20 years to Jun. Need stable purchasing power to meet those rocketing bills? The trusts' own set-asides were not always enough. Extra reserving in the owner's hands is essential: see 'Derisking' below for one method.


CONSTITUENTS AND CAPITAL
Briefly, individual trusts' contributions. First, four income metrics: compound annual dividend growth before inflation, 2002-22 (2); number of real cuts year on year in past decade; change in latest year's payout from 2020-21; months in revenue reserve at latest financial year end:

CTY: 4.6%, 2, +1.8%, 6
DIG: 3.3%, 5, +0.8%, 15
EDIN: 4.2%, 3, +5.8%, 14
IVI>AEI: 8.1%, 2, +7.1%, 10
MRCH: 2.5%, 8, +4.0%, 7
MUT: 3.9%, 8, +0.7%, 10
SCF: 4.1%, 3, +1.6%, 15
TMPL: 2.4%, 3, +7.4%, 5
--------------------------------------
B8: 4.2%, 4, +2.1%, 10


The constituents least likely to sustain real dividend rises, Merchants and Murray Income, are among those most anxious to flaunt high starting yields. City of London has been best able to arbitrate between an eye-catching yield and income growth, though its revenue reserve remains almost as low as Merchant's.

Schroder Income also has a sound all-round, long-term history and with much stronger reserving. With Edinburgh and Temple Bar the reserving must be of judgement after their mishaps; AEI has to justify inclusion after IVI was swapped out. Dunedin has gone on being costive about paying. The B8 is a unit, and it would be boring if all members behaved alike, as critics wrongly infer from the resemblances in their holdings. ITs have many more tricks up their sleeve than unit funds.


Capital metrics: average annual share price change over 10 years to Jun. 30; average yield over same period; average discount/premium; FE Trustnet Risk Score at Sep. 23:

CTY: +3.6%, 4.4%, -0.6%, 102
DIG: +1.9%, 4.5%, 8.7%, 128
EDIN: +1.4%, 4.2%, 7.4%, 131
IVI>AEI: +5.6%, 4.6%, 5.9%, 166
MRCH: +1.7%, 5.3%, 6.4%, 128
MUT: +1.9%, 4.3%, 5.7%, 119
SCF: +3.3%, 4.2%, 3.4%, 124
TMPL: +1.7%, 3.5%, 3.7%, 173
------------------------------------------------------
B8: +2.3%, 4.2%, 4.8%, 113


Discounts ranged from stingy DIG's 9% to a small premium for CTY. Averaging 5%, they have tightened from the long-run 8% or so: another sign of retail demand for income whose purchasing power is not gnawed away by inflation or the need to restock revenue reserves. A demanding mission, but no other asset type can come close for running yield.

Risk Scores are measures of share price volatility relative to the whole market, weighted towards the present and based on movements during the past three years. Cash is 0, the FTSE 100 index 100. The basket has 'risked down' a little since Jun. 2021. The proverbial Steady Eddy, CTY, is virtually a Footsie tracker. Aberdeen Equity Income is the most erratic due to its liking for small caps. All others are 20-30% more lively than the Footsie.

Capital preservation over a long holding period is an aspiration for this basket, not a firm goal. It is reasonable to hope in that if income holds its own, so should values. But preservation cannot be guaranteed over any timespan, let alone over five or ten years. Income is for income's sake. Any attempt either to bump it up, or to nail down paper gains, by tinkering voluntarily with the original, equal-weight choice invalidates the method.

The only reason to follow share price movements is to check that income is not rising too fast at capital's cost. If punters are marking down trusts because their dividends are seen as parlous, watch out.



PERFORMANCE 2000-22
Let us see how the B8 would have performed in practice. An investor places the same £75,000 lump sum as went into pyad's original High Yield Portfolio, HYP1. Same equal weighting and 1% acquisition costs. Bought on the same day: Nov. 10, 2000.

The basket would have collared £5,715 (£5,526) of income in its 22nd year to Jun. 2021, a 3.4% nominal increase. The Basket of Seven, trying more for growth of income, would have bagged £6,829. After 21.6 years the B8 would have dispensed £86,034 of dividends while the Basket of Seven earned £97,386. The tortoise overtook the hare some time back.

The Basket of Eight's yield from Jul. 2021's brought-forward capital was 4.8% against its lifetime average of 4.3%. The problem of actual collections undershooting inflation looms large presently; but this level of reward stands up well against cash or fixed interest, if the basket be viewed as a savings account.

The B8's market value shank by 3.7% to £114,345 (2021: £118,728, up 31% on the relief rally from covid). Its best at a month-end was £129,156 back in Dec. 2019. From launch until Jun. it has compounded at 2.0% pa. The cost of living rose by about 3% pa, so real capital was eroded. For relativists, the All-Share Index grew by 1.3% pa. Total return would make the All-Share look worse, since it furnishes smaller dividends.

The basket depreciated during ten of 21 years. Outperformance of the FTAS occurred eleven times out of 21. Half the B8's shortfalls, as mentioned, are recent. Overall it did modestly better than an index tracker.

In 259 monthly changes from launch to end-Jun. 2022, market value has risen or stood still 152 times at an average +3.1%, and fallen 107 times averaging -3.7%. On only 28 occasions did the basket gain 5% or more in a month, and on 21 lost as much, so the short term was normally calm sailing. Respective figures for the All-Share are 146 times on the up, month on month, averaging +2.9%; 113 going down, by -3.1% average. The B8's edge in upmoves kept it in front overall, though it has lagged the FTAS in the last ten years.



DERISKING
Added safety comes from 'derisking' the income. One mimics an index-linked bond and an income reserve backs it up. Spendable income somewhat below the raw total generated is chosen and the difference assigned to a reserve.

Using forecast receipts for the calendar year to Dec., the £75,000 basket here illustrated could have been derisked to give a 3.0% yield in its first eight months' operation as spendable income, against a historic All-Share yield of 2.2% at inception. That would have absorbed all collections in the inaugural 'stub' period (eight months to Jun. 2001); but the B8's income in its infancy could have combined inflation protection for the 3% return with a reserve growing to 12 months by Jun. 2006.

Thereupon an increase of one-fifth in the withdrawal rate to 3.9%+RPI would have been compatible with 10 months' reserve after ten years, at Jun. 2007. The buffer has been dented by KungFlu and Kyiv-- not grievously, but it would be rash to bump withdrawal higher while the consensus foresees subdued economic growth and dividend increases for several years. Leaving the 3.9%+RPI be, the reserve kitty should be ten months at Dec. This exceeds the bottom of nine months reached six years ago, and above the lifespan average's seven months, which embraces the build-up before the withdrawal rate was enhanced. On balance, a reassuring picture.

The B8's 3.9% yield for most of the its existence stands against a similarly derisked 5.2% from its inspiration, pyad's High Yield Portfolio, based on HYP1's results till last Nov. HYP1 has shelled out a far larger gross income, tempered by the need to iron out its ups and downs by setting more aside.

Derisking would have required 5% of the basket's receipts to be held back, over and above the ~3% which trusts retained. It is a hypercautious strategy for those who cannot let income's buying power wobble. Such savers might opt to accept the lower initial yield of a 'growthier' income portfolio, unless they expect to drop off the twig fairly soon. The B7 now yields 4.2%.

All B8 members shell out quarterly: dosh arrives little and often, just under once a fortnight. A cost-effective lump sum would be £10,000 or more gross. With stamp duty of 0.5% and commission of £12.50 a share, ten grand gets a starting income of £502 pa at Sep. 23's prices, averaging £15.69 for 32 dividends a year. Latest tax changes make dividend income more attractive for those who can commit much bigger amounts.
----------------------------------------------------------------------------------------------------------
[1] Review for 2021:

viewtopic.php?f=54&t=32455&p=464307#p464307


Review for 2020:
viewtopic.php?f=54&t=25362&p=405863#p405863

Review for 2019:
viewtopic.php?f=54&t=19695&p=257088&hilit=Basket+of+Eight#p257088

Latest B7 review to Mar. 2022:
viewtopic.php?f=54&t=34002&p=492277#p492277

[2] Dividends' compound annual growth rate is measured from Apr. 2001, to ignore arbitrarily different payment dates and numbers during the portfolio's first eight months.