It is impossible for a man to learn what he thinks he already knows.
Epictetus

Portfolio distributions that have been tracked in the journey so far have had two important but distinct components: investment income (such as interest or dividends), and realised capital gains.

What ultimately matters for reaching any financial independence target is total returns – which are the sum of capital gains and investment income. These two components, working together, push forward progress on the voyage. Distributions through the journey, however, provide an important and tangible measure of progress.

This longer read post explores through past portfolio data the level and significance of realised capital gains I have received in regular Vanguard retail fund distributions. It also analyses the level of ‘pure’ income – that is, counting only interest and dividends – produced by the FI portfolio and discusses what this means for managing the portfolio in the future.

Analysis of the two distinct components – capital gains and income – of my Vanguard retail fund distributions helps in understanding past and future variations in the level of these distributions, and the sustainable long-term income potential from the portfolio.

How these Vanguard distributions are structured, have behaved, and what they can be expected to do in the future is an important question for my financial independence portfolio – as by value Vanguard funds currently constitute over half its total value.

Establishing the level of distributed capital gains

Past portfolio income updates, as well reviews of past distributions and future portfolio income have noted in passing the role that distributions of capital gains realised within my Vanguard retail index funds have played in large shifts in the level of total portfolio distributions.

To get a more systematic and data-led picture of the situation, I have gone back and established through around 20 different annual tax and distribution statements the estimated make up of annual portfolio distributions, separating out realised capital gains that were distributed by Vanguard funds from other distributed income.

This separation allows a better view of the ‘pure’ income potential of the portfolio to emerge (i.e. the income produced without the sale of any assets or any capital gains being realised). Whilst it is correctly argued that dividends are conceptually identical to ‘forced sales’, equally, there is something intriguing about knowing how this ‘pure’ income measure compares to total portfolio distributions.

It is also crucial to know whether high past distributions are built on the sand of large one-off realisations of capital gains by the Vanguard funds which will not be sustained over time. Simply blithely relying on such distributions without recognising their one-off capital gain component could lead to an unintentionally high draw-down of assets.

The approach adopted in deriving this new income-focused measure is detailed further below (see explanatory note #1).

Charting the distributions and capital gains data

The chart below sets out annual distributions, separated into distributed capital gains (red bar and bolded figures) and other portfolio income (blue bar) across the life of the portfolio to date.

This chart below further illustrates the proportional breakdown of income and distributed capital gains in each year.

From these two charts, several feature are noticeable.

For much of the early part of the journey, distributed capital gains have played a minimal or limited part in absolute terms, and have represented a small proportion of total distributions. Yet, in a couple of distinct phases they have contributed substantially.

As an example, across 2006 and 2007 capital gains represented between 21 and 27 per cent of total distributions. In dollar terms, capital gains contributed around $5 000 to total distributions of $23 000 across the two years.

Since 2014, capital gains have constituted a growing and significant component of distributions, rising from 12 per cent in 2014 to a maximum of 48 per cent – or nearly half – of total distributions in 2017.

So, on the surface, as the size of the portfolio has grown, it appears the level of capital distributions has generally increased as well.

An inconstant tide – the instability of capital distributions through time

There is a wrinkle in this story, however.

While capital distributions have been consistently large in the five years from 2014, between 2009 and 2013 they were relatively minimal. A similar cluster of minimal capital distributions occurred between to 2001 to 2004. This raises the possibility that capital distributions could fall to low levels again, across a multi-year period.

Should this occur, the best predictor for a ‘base’ distributions could fall to is set out in the chart below.

This is simply a version of the first chart with estimated distributed capital gains from the Vanguard retail funds removed, that is, it shows just the underlying level of ‘pure’ income distributions.

A few observations on this can be made.

First, it retains the same basic shape and features up to 2014 as the first chart. That is, the overall pattern and level of total distributions is not influenced by Vanguard capital distributions in any significant way.

Second, and by contrast, from 2015 onwards income distributions are substantially lower – between $10 000 to $30 000 in some years. In fact, the continuous increase in total distributions between 2014 and 2017 disappears when capital distributions are removed. In its place, income distributions hover in the range of $14 000 to $16 000 in those years.

Third, the same dramatic step up occurs in pure income distributions as occurs in total portfolio distributions across 2017 and 2018. Income-only distributions approximately double compared to the five preceding years.

Yet this period (in green) – 2017 and 2018 – is not just the period covered in this record. It was also a time of substantial portfolio growth. Due to this, looking at the raw level of distributions alone may be somewhat incomplete in telling the story of underlying trends in income distributions. Another approach is warranted.

An alternative measure trialled – the record of ‘income yield’

What may provide more insight is looking at the average income yield as a percentage of the portfolio. This removes distortions from the growing absolute size of the portfolio.

The chart below sets out the income yield of the portfolio on two different bases.

Firstly, the simple income yield (blue) derived from income distributions as a proportion of average portfolio value in each year is given. A second ‘adjusted yield’ measure (red) is also tracked. Simple linear trends for both are given as dashed lines.

The adjusted yield line is based on removing from historical income distributions a sizeable stream of income received from high interest savings accounts (and more recently the Peer to Peer lending platform Ratesetter). Especially in the early years, these holdings were mostly associated with a housing purchase.

This is done to provide an illustration of the underlying levels of yields from just standard equity and bond holdings. As such it may provide a slightly better indication of the likely income capacity of the current FI portfolio, which has fewer income producing holdings.

This evidence broadly shows that income yields from the portfolio have declined.

The adjusted yield from shares and bonds has shown some volatility. It has actually marginally increased over time in trend terms (red dashed line), but generally moved above and below around three percent.

Yields below 2.0 per cent have been relatively unusual. Over the entire period average income yields are 3.5 per cent, while the adjusted yield has been only 2.6 per cent. This lower figure shows the impact of removing relatively high fixed interest returns in the early 2000s.

As a point of comparison, the average dividend yield of Australian equities over the past 40 years has been around 4 per cent.

Some care and caveats need to be applied in interpreting the numbers in the chart above. Deriving the income yield on a single average portfolio value over a year is a necessary simplification, however, it will inevitably produce inexact estimates.

The shifts in yields will also have been impacted by shifts in overall portfolio asset allocation through time (including a bond allocation above 25 per cent between 2014 and 2016). The chart below shows some of these variation just in the past thirteen years.Despite these caveats, the income yield estimates fairly closely match some others previously derived in the post Set and Drift, on the portfolios future income potential.

Currents and winds – a relationship between capital and income gains?

The other possible question looking at the data is – do higher realised capital gains and higher income payouts tend to cluster or occur together?

Bearing in mind all the caveats discussed above, the answer on the record so far appears to be equivocal.

Years with higher than average income distributions do coincide in some years with years of high distributed capital gain. In other years, however, income distributions are high while capital gains distributed in the Vanguard funds are negligible.

Overall, there does not appear to be sufficient data enabling a firm conclusion about relationships, leaving no clear answer to the question.

From observation to navigation – implications for the voyage

This analysis shows that:

• Distributed capital gains will continue to be be significant over time – With the growth in the size of the Vanguard funds over the past two decades, capital distributions have increased. This is likely to continue, albeit at a slightly slower pace recognising that new contributions are being directed into exchange traded equity funds which do not necessarily have all of the same needs to realise and distribute capital gains. (See explanatory note #2)
• Introducing volatility into distributions – Distributed gains have ranged from small amounts to over $30 000 per year and nearly half of total distributions. For other periods, distributed capital gains have been low for continuous periods up to five years.
• Reaching financial independence from ‘pure’ portfolio income would be a challenge based on the past record – Based on the pure ‘income yield’ actually experienced since 2009 of 2.8 per cent, for example, such a task implies an initial portfolio of $3.1 million with no sequence of return risk.

Considering these findings in the context of the FI portfolio, it is not clear that any specific action is required.

Rather, what has been gained is new knowledge that will help in understanding the key qualities of otherwise difficult to predict distributed capital gains each year from my Vanguard retail funds.

Concluding observations – weighing a change in course?

These large and variable gains have some disadvantages. These include obscuring an otherwise more steady and psychologically satisfying progressive growth of the passive income stream, and triggering uneven annual taxation liabilities.

Yet these are not their only effects.

The distributed gains also provide an increased flow of investable funds, enabling the redirection of capital to stay closer to the portfolio’s target allocation. They may also serve to partially deliver on future income needs in the early years of financial independence. That is, they may reduce the need to sell other assets to meet income needs.

It is also important to recognise that part of the capital gains arise from the underlying diversified retail funds needing to stay within a pre-set target allocation.

To the extent these actions help my broader portfolio to also not drift too far from the desired allocation, the realisation of these capital gains can be viewed as partially unavoidable costs which would be incurred even with any alternative arrangements. That is, were the entirety of portfolio funds in asset specific exchange traded funds, some of these capital gains would still need to be realised in any case, in a process of manual rebalancing.

For these reasons, the sale of these funds to seek to ‘avoid’ these uneven capital distributions is not an obvious or clear-cut decision. Nor is it clear that having large capital distributions is itself so disadvantageous that it outweighs some of its side benefits.

Rather, it may be that simply understanding the directions of the wind, and the currents underneath the ship, should enable safer and more confident navigation into unknown waters ahead.

References

Reserve Bank of Australia Research Discussion Paper 19-04 –  History of Australian Equities, June 2019 (pdf)

Explanatory notes

1. The data used was primarily Vanguard annual tax statements from financial years 2001-02 to 2018-19. Up to a change in approach in 2018 driven by changing tax law requirements these statements provide a clear defined value for the realised capital gains component of cash distributions. Following this change, no specific attributed dollar value linked to the cash distribution is set out, however, a ‘total capital gains’ figure is still given. This has been used from 2018. To provide the calendar year data given here from the financial year reporting it was necessary to make an assumption about the half-yearly timing of the capital gains identified. This was undertaken on the basis on a (June) 60/40 (December) split between the two half years, which appeared the closest simplification to the theoretical even split that should be present which did not result in plainly illogical outcomes (for example, attributed realised capital gains being higher than the actual cash distribution known to be paid that half-year). This means that the estimated gains attributed to any particular year are hypothecated and provisional, but likely to be correct when measured across multiple years. It also means totals of calendar year distributions used derived in this manner will not match the ordinary calendar year distribution series provided.
2. The two exchange traded funds (A200 and VAS) in the portfolio appear to have distributed realised capital gains of only $359 across 2017-18 and 2018-19. That is, they are insignificant contributors to the issue at hand so far compared to the Vanguard retail fund.