Midway along the journey of our life
I woke to find myself in a dark wood,
for I had wandered off from the straight path.

Dante, The Divine Comedy: Inferno, Canto I

This is my forty-fifth portfolio update. I complete this update monthly to check my progress against my goal.

Portfolio goal

My objective is to reach a portfolio of $2 180 000 by 1 July 2021. This would produce a real annual income of about $87 000 (in 2020 dollars).

This portfolio objective is based on an expected average real return of 3.99 per cent, or a nominal return of 6.49 per cent.

Portfolio summary

Asset allocation

Presented visually, below is a high-level view of the current asset allocation of the portfolio.

Comments

The portfolio has increased in value for the fifth consecutive month, and is starting to approach the monthly value last reached in January.

The portfolio has grown over $48 000, or 2.7 per cent this month, reflecting the strong market recovery since late March

The growth in the portfolio was broadly-based across global and Australian equities, with an increase of around 3.8 per cent. Following strong previous rises, gold holdings decreased by around 2.2 per cent, while Bitcoin continued to increase in value (by 2.5 per cent).

Combined, the value of gold and Bitcoin holdings remain at a new peak, while total equity holdings are still below their late January peak to the tune of around $50 000. The fixed income holdings of the portfolio continue to fall below the target allocation.

The expanding value of gold and Bitcoin holdings since January last year have actually had the practical effect of driving new investments into equities, since effectively for each dollar of appreciation, for example, my target allocation to equities rises by seven dollars.

New investments this month have been in the Vanguard international shares exchange-traded fund (VGS) and the Australian shares equivalent (VAS). These have been directed to bring my actual asset allocation more closely in line with the target split between Australian and global shares set out in the portfolio plan.

As the exchange traded funds such as VGS, VAS and Betashares A200 now make up nearly 30 per cent of the overall portfolio, the quarterly payments they provide have increased in magnitude and importance. Early in the journey, third quarter distributions were essentially immaterial events.

Using the same ‘median per unit’ forecast approach as recently used for half yearly forecasts would suggest a third quarter payout due at the end of September of around $6000. Due to significant announced dividend reductions across this year I am, however, currently assuming this is likely to be significantly lower, and perhaps in the vicinity of $4000 or less. 

Finding true north: approach to achieving a set asset allocation 

One of the choices facing all investors with a preferred asset allocation is how strictly the target is applied over time, and what variability is acceptable around that. There is a significant body of financial literature around that issue.

My own approach has been to seek to target the preferred asset allocation dynamically, through buying the asset class that is furthest from its target, with new portfolio contributions, and re-investment of paid out distributions.

As part of monitoring asset allocation, I also track a measure of ‘absolute’ variance, to understand at a whole of portfolio level how far it is from the desired allocation.

This is the sum of the absolute value of variances (e.g. so that being 3 per cent under target in shares, and 7 per cent over target in fixed interest will equal an absolute variance of 10 per cent under this measure).

This measure is currently sitting near its highest level in around 2 years, at 15.0 per cent, as can be seen in the chart below.

The dominant reason for this higher level of variance from target is significant appreciation in the price of gold and Bitcoin holdings.

Mapping the sources of portfolio variances

Changes in target allocations in the past makes direct comparisons problematic, but previous peaks of the variance measure matches almost perfectly past Bitcoin price movements.

For a brief period in January 2018, gold and Bitcoin combined constituted 20 per cent, or 1 in 5 dollars of the entire portfolio. Due to the growth in other equity components of the portfolio since this level has not been subsequently exceeded.

Nonetheless, it is instructive to understand that the dollar value of combined gold and Bitcoin holdings is actually up around $40 000 from that brief peak. With the larger portfolio, this now means they together make up 17.2 per cent of the total portfolio value.

Tacking into the wind of portfolio movements?

The logical question to fall out from this situation is: to what extent should this drive an active choice to sell down gold and Bitcoin until they resume their 10 per cent target allocation?

This would currently imply selling around $130 000 of gold or Bitcoin, and generating a capital gains tax liability of potentially up to $27 000. Needless to say this is not an attractive proposition. Several other considerations lead me to not make this choice:

• The problem may solve itself as portfolio grows – Growth and continued investments in the portfolio will tend to reduce the variance caused by gold and Bitcoin. The asset allocation targeting approach I adopt has seen continued contributions to equities, reducing the ability of these alternative assets to add to future variance.
• Falls in Bitcoin or gold values will also solve the problem – Conversely, price falls in Bitcoin or gold will tend to reduce the variance issue, and such price falls have significant precedents, with for example Bitcoin holdings falling to a value of around $50 000 as recently as January 2019.
• If neither of these happen, there may be bigger issues to solve – The only scenario where neither of these alleviating factors occur is should gold and Bitcoin continue to rapidly appreciate compared to other assets, in which case it is difficult to see the value of reducing exposure now.
• Does Bitcoin even fit the asset allocation model? – Bitcoin in particular is not a well established or accepted asset class as yet, so it may not be appropriate to apply traditional allocation rules to it – it may be functioning more as a hedge or option against extreme states of the world. Linked to this is the high degree of volatility in Bitcoin. Adopting too tight a target on Bitcoin holdings would potentially see a need to buy and sell Bitcoin frequently, where my intention is to actually never purchase any more.

This approach is a departure from a mechanistic implementation of an asset allocation rule. Rather, the approach I take is pragmatic.

Tracking course drift in the portfolio components

As an example, I regularly review whether a significant fall in Bitcoin prices to its recent lows would alter my monthly decision on where to direct new investments. So far it does not, and the ‘signal’ continues to be to buy new equities.

Another tool I use is a monthly measurement of the absolute dollar variance of Australian and global shares, as well as fixed interest, from their ideal target allocations.

The chart below sets this out for the period since January 2019. A positive value effectively represents an over-allocation to a sector, a negative value, an under-allocation compared to target.

This reinforces the overall story that, as gold and Bitcoin have grown in value, there emerges a larger ‘deficit’ to the target. Falls in equities markets across February and March also produce visibly larger ‘dollar gaps’ to the target allocation.

This graph enables a tracking of the impact of portfolio gains or losses, and volatility, and a better understanding of the practical task of returning to target allocations. Runaway lines in either direction would be evidence that current approaches for returning to targets were unworkable, but so far this does not appear to be the case.

A crossing over: a credit card FI milestone

This month has seen a long awaited milestone reached.

Calculated on a past three year average, portfolio distributions now entirely meet monthly credit card expenses. This means that every credit card purchase – each shopping trip or online purchase – is effectively paid for by average portfolio distributions.

At the start of this journey, distributions were only equivalent to around 40 per cent of credit card expenses. As time has progressed distributions have increased to cover a larger and larger proportion of card expenses.

Most recently, with COVID-19 related restrictions having pushed card expenditure down further, the remaining gap to this ‘Credit Card FI’ target has closed.

Looked at on an un-smoothed basis, expenditures on the credit card have continued to be slightly lower than average across the past month. The below chart details the extent to which portfolio distributions (red) cover estimated total expenses (green), measured month to month.

Credit card expenditure makes up around 80 per cent of total spending, so this is not a milestone that makes paid work irrelevant or optional. Similarly, if spending rises as various travel and other restrictions ease, it is possible that this position could be temporary.

Equally, should distributions fall dramatically below long term averages in the year ahead, this could result in average distributions falling faster than average monthly card expenditure. Even without this, on a three year average basis, monthly distributions will decline as high distributions received in the second half of 2017 slowly fall out of the estimation sample.

For the moment, however, a slim margin exists. Distributions are $13 per month above average monthly credit card bills. This feels like a substantial achievement to note, as one unlooked for at the outset of the journey.

Progress

Progress against the objective, and the additional measures I have reached is set out below.

Summary

What feels like a long winter is just passed. The cold days and weeks have felt repetitive and dominated by a pervasive sense of uncertainty. Yet through this time, this wandering off, the portfolio has moved quite steadily back towards it previous highs. That it is even approaching them in the course of just a few months is unexpected.

What this obscures is the different components of growth driving this outcome. The portfolio that is recovering, like the index it follows, is changing in its underlying composition. This can be seen most starkly in the high levels of variance from the target portfolio sought discussed above.

It is equally true, however, of individual components such as international equity holdings. In the case of the United States the overall index performance has been driven by share price growth in just a few information technology giants. Gold and Bitcoin have emerged from the shadows of the portfolio to an unintended leading role in portfolio growth since early 2019.

This month I have enjoyed reading the Chapter by Chapter release of the Aussie FIRE e-book coordinated by Pearler. I’ve also been reading posts from some newer Australian financial independence bloggers, Two to Fire, FIRE Down Under, and Chasing FIRE Down Under.

In podcasts, I have enjoyed the Mad Fientist’s update on his fourth year of financial freedom, and Pat and Dave’s FIRE and Chill episodes, including an excellent one on market timing fallacies.

The ASX Australian Investor Study 2020 has also been released – setting out some broader trends in recent Australian investment markets, and containing a snapshot of the holdings, approaches and views of everyday investors. This contained many intriguing findings, such as the median investment portfolio ($130 000), its most frequent components (direct Australian shares), and how frequently portfolios are usually checked – with 61 per cent of investors checking their portfolios at least once a month.

This is my own approach also. Monthly assessments allow me to gauge and reflect on how I or elements of the portfolio may have wandered off the straight way in the middle of the journey. Without this, the risk is that dark woods and bent pathways beckon.