Wind in the Sails – A History of Portfolio Distributions

IMG_20170526_120759_641
Gradually, and then suddenly.
Hemingway The Sun Also Rises

The wind stirs and grows

In this journey portfolio distributions have represented an extra wind in the sails, whose contribution should not be underestimated. In fact, total distributions over the past 19 years total over $360 000, or more than 25 per cent of the total current value of the portfolio.

When I started tracking portfolio income, distributions represented just a weak breath of wind. They made a tiny monthly contribution of $27 or so. This has since turned into a strong gust, with over one-third of the total value of all distributions being earned in the past two financial years.

By way of contrast to this recent strength, it took 13 long years of saving and investing to earn the first one-third of total distributions received. Similarly, it took 11 years of saving and investing for portfolio distributions to break decisively above around $10 000 per year. The below figure sets out past recorded distributions in 2017 dollars.

Port distThe contributions of different investments have changed over time. A significant part of very early income was actually linked to cash holdings in high interest savings accounts.

Through a process of gradual investment in Vanguard’s high growth indexed fund, the distributions from this source have become dominant. This source now drives overall distributions performance, save for some growth in Australian shares ETFs in the past year or so.

It is noticeable that there is variability, including peaks and falls. In some years capital gains were realised by funds and paid out, meaning peaks in what are distributed gains that go beyond income or dividends. An example of this is achieving nearly $20 000 of distributions in 2005-06, a level not exceeded again until five years later in 2010-11, which itself was a peak not exceeded for another five years.

Measuring the wind – the rate of portfolio distributions

To help forecast the level of future portfolio distributions and track progress to my goals, I have constructed estimates of the portfolio income rate of the past two decades. This is calculated by determining the total distributed income over the financial year by all portfolio assets, and dividing it by the average portfolio level half way through the year. Note that where large movements occur unevenly through a year, some minor inaccuracy is introduced.

The theory is, instead of seeking to project forward a trend in the absolute level of distributions, why not seek to observe what the historical level of portfolio income has been produced, based on the average of the total portfolio.

Average port distOver ten years, the median level of distributions from the portfolio has been 4.4 per cent. The mean average has been higher, at 5.1 per cent. This now allows me to have some degree of certainty around the likely bounds of future distributions from any projected portfolio level.

There are some anomalies in the figures, caused by things such as a major house purchase changing the size of the portfolio, and the adoption and abandonment of a range of different financial products. Through the period, also, interest rates have been steadily falling.

One factor that does not seem to have been a particular driver of variability is asset allocation. This is perhaps surprising as over the past decade non-income producing holdings (gold and Bitcoin) have been introduced and started to formed a small part of the portfolio – generally 5 to 10 per cent.

Although in general the portfolio has moved to very low cash levels and a reduced level of fixed interest products through time, overall equity allocation has remained within a few percentage points of 60 per cent since 2007.

Fair winds and following seas?

The past two years have seen slightly higher than usual distributions levels. It remains to be seen whether these are temporary anomalies. Similarly, the absolute level of portfolio distributions in the past two years has been decisively highly than historical levels.

This leads to mindfulness of the potential for future reverses in the absolute level. Nonetheless, applying the historical 4.4 per cent average to my current portfolio level still produces a forecast annual distribution income of around $62 000, above the first of my current targets.

Such projections can’t protect against storms ahead, but still provide a comforting thought on the continuing journey.

6 comments

    1. Thanks for the comment. Yes, the stability, particularly around the median of 4.4% surprised me. I guess time will tell, as portfolio values going down also contributes to that stability, right?

  1. Hi. When calculating your distributions what year do you include them in for the vanguard funds? (For example, the retail funds often declare their divs on 30th June but actually pay them on the 1st July, being the new financial year). Or do you always go by calendar year?

    1. Hi Wildgoose, thanks for the read! What I do is record in financial years, but attribute the Vanguard distribution to the financial year in which it was ‘earned’, if that makes sense. So a distribution formally paid on 1 July would be recorded as portfolio income in the financial year that ended 30 June. That’s just the way I started recording it, so I’ve stuck with it, and I think there’s a logic to it.

  2. Thanks very much for sharing, very interesting. Do the dollar and yield amounts include grossing up for franking credits?

    Just eyeballing the graph it looks like the mean has been dragged up a fair bit by a few big outliers, if you remove those presumably the mean would be a lot closer to the 4% level although it might not change the median a great deal? I seem to recall you’re targeting an income level of $80k pa eventually, would that be based on actual distributions at that time or would you want it to be $80k pa assuming that average or median level, or to be super conservative based on your lowest distribution yield? Sorry, a lot of questions there!

    Also one last thing, you said a median of 4.4% and average of 5.1% distributions under the graph but then use 4.4% as the average in the second last paragraph? I

    1. Hi Aussie HIFIRE – good questions!

      On the grossing up, I have tended to record it all as ‘cash that was received’, certainly for the major Vanguard funds and ETFs. I may have, I’d need to go back and check, inconsistently added a few small share distributions in at their grossed up value, but they would be 2-3% of my portfolio so this wouldn’t have much on an effect. But your broader point is right, the 4.4% is not counting franking credits.

      Yes, correct, the outliers are the reason I prefer to think about the median value. I call that an average too (your last point).

      That’s right, my objective #2 is $80 000. per annum. At the moment that’s based on an assumed long-term real return of 3.92% on a lump sum of 2.04 million. That return level is based on long-term returns from an assumed asset allocation.

      So it’s not strictly speaking formally connected to what I am recording here. The reason for that is that to some degree past distributions could reflect capital gains payouts from time to time, and be systematically higher than longer term returns for some time, and I wouldn’t want to base a FI decision on that, i.e. if I was taking out too much capital and risking under-saving. I have been thinking about this issue a lot, as there are obvious intuitive arguments the other way (if distributions routinely meet expenses, that’s one level or definition of FI).

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.