The Conviction Five: 2006-22 (closing)

[Luniversal]

In Dec. 2012, at The Motley Fool, I proposed a portfolio called the Conviction Five. It was designed, to quote one of its members' professed aims, to preserve wealth and to grow it 'in that order'. Not losing much even in crashes was paramount. It is on the Growth Strategies board faute de mieux.

Background and characteristics here:

viewtopic.php?f=96&t=32752&p=469844#p469844

The chosen Five were Capital Gearing (CGT), Independent (IIT), Lindsell Train (LTI), Personal Assets (PNL) and Ruffer (RICA), which last is not a British authorised trust but a Guernsey 'investment company' listed in London (1). All are midsized, dedicated more or less firmly to not losing money. They are tended by opinionated maestri relaxed about acting athwart conventional wisdom.

The C5 is losing a member: Independent is being taken over, for shares or cash, by Monks on the retiral of Max Ward, the managing director throughout its 22-year life. Monks is not exactly a conviction trust, though as a global-growther it sticks its neck out on high tech. I might have preserved the C5 brief by buying RIT Capital Partners (RIT) as a replacement. But between them my 'Defensive Three' (D3) portfolio for wealth preservation (2) and the 'Growth Ten' (G10) for capital accumulation cover the Conviction Five's ground. The G10 includes Monks; the D3 incorporates RIT. So I am calling time on the C5.

This closing account values IIT at the recent market price of 227.5p a share (3), plus a pre-liquidation dividend of 9p payable on Nov. 4. Results have been calculated back to that fate-tempting Friday the Thirteenth in Jan. 2006... as with my paper High Yield Portfolio, HYP06, and my other investment trust collections: the D3, G10 and the income-focused Baskets of Seven (B7) and Eight (B8).

All these assume £75,000 is invested before 1% initial costs, allocated equally to their constituents. They operate on autopilot except when events such as a takeover compel intervention: no tinkering or ploughing-back of income. Deflated figures use the Retail Prices Index; Sep. 2022's inflation, out next week, is taken as 12.3%, unchanged from Aug.

The model portfolios' objectives differ starkly. To see how widely they diverged in combined return, and how it splits between bird-in-hand dividends and uncashed capital gains, tells us about the right horse for one's course. It also reveals volatility along the 17-year line of march. The period took in a brief spell of calm, a dire, compacted slump in equities (2007-09) and a very long but slow and fitful recovery, halted temporarily by a weird, sharp but brief pan(dem)ic which have way to a spasmodic rally. Such conditions challenged the C5's wealth-conservationist powers.


CAPITAL

The Conviction Five's Dec. 31 (2022: Sep. 30) market values; nominal and real percentage changes/performance in percentage points plus or minus versus the FT All-Share Index:

2006: £84,060, +12.1, +7.7/+0.8
2007: £76,957, -8.4, -12.4/-10.5
2008: £70,932, -7.8, -5.7/+25.0
2009: £88,518, +24.8, +22.4/-0.2
2010: £104,493, +18.0, +17.2/+7.1
2011: £104,743, +0.2, -4.6/+6.9
2012: £116,310, +11.0, +7.9/+2.8
2013: £124,875, +7.4, +4.7/-9.3
2014: £131,970, +5.7, +4.1/+7.8
2015: £159,806, +21.1, +19.9/+23.6
2016: £208,661, +30.6, +28.1/+18.1
2017: £219,673, +5.3, +1.2/-3.7
2018: £243,389, +10.8, +8.1/+23.7
2019: £259,609, +6.7, +4.5/-7.5
2020: £294,612, +13.5, +12.6/+25.9
2021: £286.457, -2.8, -10.3/-17.3
2022 (Sep. 30): £237,460.-19.4, -31.7/-8.8


The C5 increased its deflated value in twelve out of 17 years, gaining in all years after 2008 except the last two, which was the only time it underperformed in consecutive periods. It is quitting on a downbeat, though the good times were much better.

The portfolio would have beaten the index in ten of 17 years. To maintain initial purchasing power it needed to grow from £75,000 to £138,000 by Sep. Actual growth was nearly a hundred grand more. The C5's compound annual growth rate (CAGR) was 7.0% pa or 3.5% after inflation.

Nominal-value CAGRs per trust were: CGT, 5.1%; IIT, 3.5%; LTI, 12.7%; PNL, 3.7%; RICA, 4.9%. All beat inflation except Independent, which matched it.

Resilience already showed in the early days. The Global Financial Crisis poleaxed share prices from 2007. The portfolio would have fallen by about 8% then and again in 2008, but shrugged off the crisis by regaining one-quarter in 2009: in real terms; it recaptured lost ground since end-2006. There was a further advance of 18% in 2010.

Afterwards the C5 would have ridden the gentler upswing which became the Quantitative Easing silver age for asset prices, their longest modern bull run. The portfolio was higher in nominal worth at ten New Year's Eves from 2011 to 2020, and by at least 5% nine times in a row until things got chillier and choppier in spring 2020.

Not that the quintet always matched the All-Share Index: it lagged in 2013, 2017 and 2019. In 2021 it was 17 points behind the benchmark, its worst yet. The subsequent nine months produced a 9-point underperformance. That is relative: as a partly growth-focused enterprise, the C5 is more aptly stood against the Growth Ten, whose coverage ought to reflect the capitalist world's vivacity more than FTAS businesses alone. Did it?

Since inception, the G10's actual value has been as likely to grow each twelvemonth. It exceeded the convicts in 2007, finishing £8,000 ahead. After the V-shaped crash and rapid rally, at end-2009 the C5 was £16,000 ahead. It held the lead until 2020, but by Dec. 2021 it was running £7,000 behind. After three-quarters of 2022, however, the G10 was almost £19,000 behind at a market value of £216.806. Foreign growth stocks, contemplating a low-growth world, have gone out of fashion despite sterling's slide against the dollar.

After the GFC the Conviction Five unearthed a megastar in Lindsell Train, but the Growth Ten harboured the still more sensational Scottish Mortgage Trust (SMT). They went up near-vertically together and descended precipitately together-- LTI less so, as befitted the C5.



CONTRIBUTORS

The C5 has evolved as a kind of barbell strategy. At one end, Lindsell Train and Independent have been highly volatile, though tending to grow fastest; at the other, Ruffer, Capital Gearing and above all Personal Assets have been safety-first stabilisers at the expense of growth.

Let us consider the year-end percentage weights at six moments: 2007, when the market was beginning to crumble; 2009, after the QE medicine; 2013, following a lesser fit of punters' nerves; 2016, when the charge into Lindsell Train began; 2020 and 2022, after covid fright and relief, and after LTI had blown off:

CGT: 19.1/21.4/18.4/13.0/11.5/14.7
IIT: 17.8/12.6/13.5/11.6/11.1/11.2
LTI: 24.3/23.2/32.7/50.7/57.6/48.1
PNL: 19.7/18.8/15.3/11.2/9.2/11.7
RICA: 19.1/24.0/20.2/13.5/10.6/14.3


Constituents have not altered their mandates or methods since the pandemic. Two bulletins have tracked them through the turbulence:

2021 report:

viewtopic.php?f=96&t=32752

2020 report:
viewtopic.php?f=96&t=27375

For better and then worse, Lindsell Train dominated the collection in the last third of its history. Latterly a widespread suspicion arose that those wonder boys of Deep Value have lost their grip on megatrends. They did not embrace Baillie Gifford-type tech whizzery, though that abstention has been vindicated. Train and Lindsell were held to be too fond of Japan rather than China; too fixated on venerable 'moat' businesses and luxury brands as a defence against recrudescent inflation; too hesitant about ditching dogs such as Pearson. LTI's share price, £1.28 in Jan. 2006, sank from £14.63 in Dec. 2020 to £9.65 by Sep. 30.

Its adherents are staunch: there has been no fugue of the footloose such as Invesco suffered when Mark Barnett lost the plot. The trust's, and the C5's, reliance on LTI's effectively controlling minority stake in the LT fund management business has been conservatively valued, but it zoomed as the firm expanded, primarily in OEIC sales. That made the C5 markedly less 'pure' as a fund of funds, but it is difficult to carp considering how it supercharged both capital and income.


METRICS

The weighted aggregate of financial results for the composite year to Jun., which best fits different accounting dates, discloses these measures of the C5's robustness: percentage change in deflated net asset value per share; year-end (discount) or premium/expenses as percentage of year-end net asset value:

2006: 8.2, 4.9/1.01
2007: 7.9, 0.8/0.95
2008: -5.3, 2.7/1.01
2009: -9.1, (0.5)/1.11
2010: 23.5, 7.0/1.05
2011: 4.2, 2.4/1.03
2012: 3.8, 3.9/0.95
2013: 9.7, 3.7/1.06
2014: -0.6, 4.2/0.97
2015: 13.3, 3.1/1.03
2016: 6.0, 15.6/0.92
2017: 18.2, 22.2/1.02
2018: 11.1, 24.4/0.94
2019: 11.4, 43.2/0.89
2020: 4.2, 7.7/0.71
2021: 17.1, 13.9/0.72
2022: -11.7, 0.2/0.60

Same metrics by trust, average 2006-22:

CGT: 3.6, 4.5/1.05
IIT: 7.6, (8.1)/0.33
LTI: 11.8, 15.4/1.83
PNL: 1.7, 1.0/0.88
RICA: 2.7, 1.3/1.28
--------------------------
C5: 6.6, 9.4/0.94


Real NAV per share has risen by 6.6% pa and in thirteen of 17 years. The only serious setbacks for share prices were 2009, when they fell by 12.0%, 2020, down 22.1%, and in the final nine months, down 23.5%. Mr Market exacted a purchase premium continually except in 2009, when optimism temporarily left safety shots in the locker; discount control has tightened for the three safety-first members, making the portfolio average near-par betwixt NAVs and prices in 2021-22. Throughout the C5's run Independent alone generally traded at a discount, LTI at a steep premium.

Expenses average 0.94% of NAV; they too peaked in 2009, but have tended to be reduce, although the elephantiasis at LTI cut the average OCR from 0.89% in 2019 to 0.60% in the latest period. The downtrend is typical of my models, indicating good board oversight and managerial response, with performance fees abolished.

Some people shrink from paying more than asset value for trusts. The C5 has sold for about a tenth more during its life. In 2020 disillusionment with LTI clipped that bagger from a barmy 43% premium to under 8%. It now stands at 6.1%. The quotes of LTI and CGT are very heavy, rendering small orders and recycling of income awkward. Personal Assets felt our pain and did a 100:1 split in Jul. The other two seem to use a heavy price to deter investors of slender means.


VOLATILITY

The FE Trustnet Risk Score (measuring volatility in prices over the last three years, weighted towards the present), was 121 at Oct. 9 where cash is 0 and the FTSE 100 index 100. Last Dec. it was 119 and a year earlier 115. The C5 remained quiet while turbulence worsened.

The portfolio's market value compounded at just below 7% pa real across the last ten financial years. Given the relatively low volatility compared with other types of specialist equity portfolio, this is not bad going, amounting to rather more than wealth preservation. And that is before considering income.

For constituents, ten-year CAGRs of share prices (% pa), standard deviations*, and Risk Scores at Oct. 6 are:

CGT: 5.5/44/59
IIT: 11.3/203/187
LTI: 16.0/138/216
PNL: 4.0/36/50
RICA: 4.2/66/83
---------------------
C5: 10.3/14/121
*Of year/year price changes, 2012-22


The gap between SDs for the two more venturesome members and the timid trio stands out a mile. Independent again failed to excuse its jumpiness by a strong performance as a stock rotator, and Lindsell Train's when separated from the fund management side has not been effulgent since early 2020. Hence I dissolve the C5 with few regrets. Maybe its best days predated the WuFlu. and Independent's surrender is a belated instance of pyad's 'market trading': the hidden hand that succeeds better than tinkering.


INCOME

Dividends, I was wont to say, are 'best regarded as a sweetener.' In retrospect that was too dismissive. Thanks to LTI, the C5 became more juicy. Its revenue receipts somewhat alleviate the capital's recent frailties.

From 2006 the historic yield averaged c. 1.5%, against c. 2% for the G10, 3.6% for the All-Share Index, 3.5% for the Basket of Seven and 4.5% for the Basket of Eight. No C5 trust professed to seek ever-rising income. Nor have they promised to resist passes, cuts or freezes. Such policies would compromise the overriding necessity to skirt loss of value: they would nudge managers to forage among higher-yielding shares, which might well incur markdowns. The more defensive C5 members' leanings towards index-linked and other fixed interest securities, cash, gold bullion, growth stocks etc militate against a so-called progressive dividend policy.

However trusts legally have to fork out 85% of their revenue. Outcomes tell a rather different tale from theory. Here are nominal and real percentage changes between calendar years and purchasing power/indexed to 2006:

2006: £921/100
2007: £1,113, +20.9, +16.9/117
2008: £1,232, +10.7, +9.8/128
2009: £1,564, +26.9, +24.5/160
2010: £1,726,+10.3, +5.5/169
2011: £1,516, -12.2, -17.0/140
2012: £1,517, +0.1, -3.0/136
2013: £1,966, +29.6, +26.9/172
2014: £1,973, +0.4, -1.2/170
2015: £2,115, +7.2, +6.0/180
2016: £2,687, +27.0, +24.5/225
2017: £2,785, +3.6, -0.5/224
2018: £3,750, +34.7, +32.0/295
2019: £4,885, +30.3, +28.1/378
2020: £6,756, +38.3, +37.1/519
2021: £7,226, +7.0, -0.5/516
2022 (nine months): £7,911. +9.5, -2.8 forecast/501


Total £51, 642 v. £43,604 for the Growth Ten since Jan. 2006. The Conviction Five's receipts grew at 13.3% pa: almost twice as brisk as its capital growth rate and more than twice as fast as the G10's income's CAGR of 6.2%.

Revenue has swelled more lustily than receipts from the baskets or my HYP-othetical. Real falls occurred in five of 16 years, but purchasing power was always at least one-sixth higher than in Year One-- usually much higher. Over the closing nine months to Sep. (little is payable in the autumn quarter) real income, if slightly crimped by inflation since early 2020, is five times as high as Year One's. All this is largely Lindsell Train's doing, but who is complaining?

Income per trust: total to cessation, percentage share of total, CAGRs (%):

CGT: £2,974, 5.8, 8.1
IIT: £8,040, 15.6, 8.5
LTI: £30,615, 59.3, 24.2
PNL: £5,379, 10.4, 2.3
RICA: £4,635, 9.0, 2.6
-------------------------------------------------
C5 total/average: £51,632, 100.0, 13.3

Personal Assets pegged its dividend back in 2013 but bestowed an unexpected extra one this summer, worth £83. Independent's terminal distribution added £562 to its tally, the second largest-- that barbell again. Capital Gearing has never cared about dividends and Ruffer's are incidental to its tireless, anxious reshuffling of assets.


CONCLUSION

Since the pandemic selloff began, at the end of Feb. 2020, the Defensive Three has netted a gain of 4.0%, the C5 4.8% and the G10 11.8%. The All-Share Index is unchanged.

The D3 is planned as a more stable repository for wealth than the Conviction Five, so let us look briefly at how it fared from Jan. 2006. On the £75,000 invested the D3 gathered £23,436 of income versus the C5's £51,642 and the G10's £43,604. The D3's market value at Sep. 30 was £157,075, against the G10's £216,806 and the C5's £237,460. Therefore combined returns, receipts plus paper gains, for these portfolios were D3 £105,814; G10, £185,310; C5, 214,101.

Growth and semi-growth triumphed, but when we stand the results against volatility, the reverse applies. The latest blended risk score for the G10 is 138, for the C5 121 and for the D3 87. It is moving less violently than the broad London equity indices, while creeping up with a real increase of 14% in the principal. about 1 point a year. And it yielded an average 1.8%: only half the FTAS yield, but a not negligible contributor, almost one-fifth of the combined return.

My hunch is that markets will be more acrobatic for the next few years. Us oldsters and snowflakes will settle for such an insomnia cure. Or the D3 could combine with a growth portfolio to limit downsides.
--------------------------------------------------------------------------------------------------------------------------------
(1) The pool from which these were fished included AVI Global Trust (AGT), formerly British Empire (BTEM); Caledonia Investments (CLDN); Hansa Investment (HAN), formerly Hansa Trust; Manchester & London (MNL); and RIT Capital Partners (RCP), which was a tossup with Independent for C5 membership.

Alternatives to conviction trusts:
viewtopic.php?f=96&t=27501&p=380672#p380672

A £15,000 position in RIT would have become worth £31,570 by Sep. 30. It paid out £5,709 of with a CAGR of 12.0%, after adopting more liberal distributions in 2012. In short, a decent also-ran far behind Lindsell Train, and another calm constituent, right for the D3. Its Risk Score is 147.

(2) Capital Gearing, Personal Assets and RIT. See

viewtopic.php?f=8&t=33269&p=479360#p479360

This explained how the C5 became a 'hero in error'-- and foreshadowed its doom.

(3) Cash realisations will be 98% of IIT's 'formula asset value', tba.

[MDW1954]

A great post, Luni. But why not learn to master tables? The tools exist to make it easy.

MDW1954

[kiloran]

A great post, Luni. But why not learn to master tables? The tools exist to make it easy.

MDW1954

Perhaps a pointer to one of the tools may be useful
http://lemonfoolfinancialsoftware.weebl ... ormat.html

--kiloran

[littledavesab]

Very useful and good work Luniversal