No man ever steps into the same river twice.

Heraclitus

The achievement of financial independence is often, correctly, framed as getting a few key principles and habits in place, and then repeating these consistently through time. This history of portfolio change steps beyond this valuable and instructive perspective.

This is because an exclusive focus on consistency of action conceals another fact of the journey – that constant principles and habits do not avoid change through the experience. In fact, change has been a continuous marker through the financial independence journey so far.

As the portfolio and its characteristics change, different experiences, issues and challenges emerge.

This post examines some of the major areas of change. It particularly focuses on changes since the commencement of this record in 2017, which covers the second half of the journey.

While each financial independence journey and portfolio is different, this is intended to highlight a few of the key changes I have experienced in building and managing the portfolio, for any interest and insights it offers others.

History of change in the composition of the portfolio

The most significant change – aside from the growth in the overall portfolio level – has been to the composition of the portfolio. That is, balance of actual investments held in different investment vehicles.

For most of the early part of the journey, the set of Vanguard retail funds (mainly the High Growth, Growth, Balanced funds) formed the core of the portfolio. In 2007, for example, Vanguard retail funds made up no less than 95 per cent of the total portfolio.

As the recorded journey started in January 2017, this legacy was still apparent. At this time Vanguard funds made up around 80 per cent of the portfolio, as can be seen below.

Some small gold, Bitcoin, and peer-to-peer lending had been added to the portfolio by 2017. Yet these were minor elements compared to the legacy retail funds that also received regular new investments.

The equivalent chart of the composition of the portfolio today is below.

The most significant changes since 2017 have been:

• New investments in exchange traded funds – with these now constituting around 31 per cent of total portfolio assets after the first equity exchange traded fund purchases in July 2017, and switching to them as the main tools for new investments from May 2018.
• The shrinking role of the Vanguard High Growth retail fund in the portfolio – with this falling from 59 per cent of the total portfolio in January 2017 to around 33 per cent today. While still important, this fund is now increasingly just a contributor to the portfolio, rather than its mainstay.
• The growth of Bitcoin – to a position where it currently makes up around 20 per cent of the portfolio by value, having in some months reached as high as 31 per cent of portfolio.

While the overall number of portfolio investment vehicles has grown over time, due to experimentation and use of new products, generally a few larger vehicles increasingly drive portfolio outcomes. Investments outside of these make relatively marginal contributions.

Changing distributions: from a breath to filling the sails

A second major change across the journey has been in the role and level of distributions. In large part, this arises simply from the sustained growth in the overall portfolio.

The changing level and shape of distributions on the journey

In the earliest part of the journey, distributions were a trivial, almost unnoticeable, part of the investment return.

As an example, in the year 2000, they were around $609, or approximately $50 per month. By way of illustration, this was around enough to meet a (then) expensive mobile phone plan. They arose in large part from interest in an ING cash savings account earning around 4 per cent, and small dividends from shares and a Colonial First State Global Equity fund.

The commencement of the record coincided with a step-change in the level of distributions received.

Following a relatively steady growth in the absolute level of distributions, 2017 saw a doubling over previous levels. From 2017 onwards, the level of distributions started consistently reaching the level of an additional full-time salary.

There is a clear exponential trend apparent in the shape of the portfolio income, aligned to the growth of the portfolio.

As an example of what this means, over 55 per cent of all portfolio distributions ever received on the journey have occurred in the 4.5 years since starting this record in 2017. This is depicted in the green bars in the chart above.

Despite this upward trend, year to year shifts in levels of distributions are also apparent. This illustrates another area of subtly changing portfolio dynamics.

Increasingly over time, relatively small variations in dividend payouts from equities – effectively flowing from decisions taken by boards and management of Australian and foreign listed companies – will have a larger impact on the level of realised portfolio distributions than any personal decisions around additional amounts invested, or re-invested.

An unfolding map: impacts of the growth in distributions

This substantial change in the level of distributions through time has had a few different effects.

First and most importantly, it has meant it has become possible to see how distributions could support a lifestyle approximating my current level of spending.

Indeed, it has become possible to conceptualise the distributions as representing a ‘wage’ from the portfolio – one which has paid around $27 per normal working hour since 2013. The chart below sets out the notional percentage of total current expenses able to be met from distributions received, on a monthly basis, over the the past 7 years.

Second, the portfolio distributions have become sufficiently large to make their re-investment a more consequential decision.

That is, the injection of distributions back into ETF purchases, for instance, has started to form a material part of portfolio momentum and growth. Additionally, the distributions have also become large enough to allow shifts in asset allocation to be effected through deliberate choices around directing distributions. This is in contrast to a previous simple policy of automatic re-investment.

Now, rather than automatic re-investment, larger half-yearly distributions are gradually fed into the portfolio over the subsequent six months, to support the asset allocation target. Smaller quarterly distributions in April and September are usually re-invested on a ‘when received’ basis.

Third, it has necessitated a system of setting aside a portion of distributions for future tax liabilities.

The calibration of this scheme is still being refined. An allowance of 25 per cent of distributions paid is my current rule of thumb for future liabilities. Obviously a circumstance which is desirable to avoid is being forced to sell valued assets at an unfavourable time just to meet unexpected tax liabilities. So far this has been avoided.

Growth in franking credits: gradually, and then suddenly

Connected to this issue of tax liabilities is the emergence of franking credits – or essentially credits for pre-payment of personal income tax – as a significant stand-alone consideration.

Franking credits were essentially an inconsequential consideration in early parts of the journey, barely significant in their own terms, or in meeting tax liabilities.

Again, the growth of the portfolio has changed this circumstance. A further factor has been a marked shift in investments towards Australian equity ETFs over the past three years, which generate significant franking credits. This has had the result of substantially increasing the role of franking credits in meeting portfolio-related tax payments.

The figure below demonstrates that over the two previous financial years franking credits have contributed more than $8,000 to total realised portfolio returns, almost doubling the next highest value for financial year 2017-18.

This means that these credits alone are likely to represent an additional offset to tax liabilities equal to around 13 per cent of projected total distributions of around $60,000 for 2020-21.

From forecasts last month, the set of upcoming distributions which are due in July appear likely to continue these trends, and the separate regular portfolio income reports will continue to provide more analysis of longer-term directions in portfolio distribution levels and composition.

Assessing changes in risk and volatility in the portfolio

The portfolio’s increasing size – together with shifts in its composition – has also led to significant changes in its experienced and potential future volatility.

This is most evident in the pattern of monthly movements in portfolio value through the history of the record in the chart below.

This chart highlights the increased dispersion of absolute portfolio movements, positive and negative, over the past four years.

Tracing the sources of increased volatility: growth, Bitcoin and markets

The portfolio has grown in size by two and a half times in the period covered in the chart above.

Absent any other change this would naturally increase the magnitude of expected monthly fluctuations in the portfolio by the same amount. The rising level of the portfolio affecting volatility in dollar terms has of course mathematically been a factor since the earliest commencement of portfolio investments.

Due to the compounding nature of the journey, however, it has come into focus only more recently. Put simply, natural and average movements in markets can now produce higher levels of fluctuation, or ‘noise’, measured in simple dollar terms than ever before.

What has added to this, however, is the the growth in the value of Bitcoin holdings. This has mostly occurred in the last 2.5 years.

From making up around one per cent of the portfolio in early 2017, through price appreciation since Bitcoin currently represents around 20 per cent of the total portfolio.

Bitcoin is one of the most volatile assets that can be held, its volatility leaving other traditional investment classes in its wake. Thus as the absolute dollar value of Bitcoin holdings has appreciated (itself representing upward volatility), the relative capacity of Bitcoin to affect the overall volatility of the portfolio is also increased.

The final remaining source of volatility is the day-to-day movements of asset markets themselves. In particular, as the portfolio grows the nominal dollar value exposure to such market movements obviously increases.

Measuring volatility in the portfolio: deviations from the course

The graph below examines the rolling 12-month standard deviation of the portfolio, over the past 30 months.

It reflects all three of these factors discussed – growing portfolio size, the impact of increased Bitcoin exposure, and market movements over that time.

In the chart the blue line tracks the rolling average standard deviation of monthly movements in the portfolio. The red line represents the same standard deviation measure, but expressed as a percentage of the monthly portfolio value.

The way to consider these are as two measures where approximately two-thirds of the time, the experienced level of portfolio volatility is equal or less than the level indicated by the lines.

What the chart shows is that:

• There is a clear ‘before’ and an ‘after’ March 2020 story of volatility – generally, in the lead up to the market events of March 2020 associated with COVID-19 and global lockdowns, most months saw variations in portfolio value of around $40,000 or less – or between 2 and 3 per cent of the portfolio value.
• After March 2020 portfolio peaked – the March 2020 period saw a doubling in volatility, leading to rolling average monthly volatility of approximately $90,000, or about 5 per cent of the portfolio value.
• On some measures volatility is slowly reducing – as equity markets have recovered and advanced, and Bitcoin has simultaneously increased in value, portfolio volatility as a percentage of the portfolio has gradually reduced, but not returned to pre-March 2020 levels.
• Dollar volatility in the portfolio, however, has not reduced – the increase in the value of Bitcoin holdings, and sustained price volatility since October 2020, has led to volatility in nominal dollar terms continuing to generally increase, to around $120,000 per month.

In turn this may suggest that while volatility in the overall level of the portfolio might gradually reduce with more stable conditions, the portfolio is likely to continue experiencing relatively high volatility in dollar terms, simply from its larger size.

Without a radical revision to asset allocation inconsistent with the overall portfolio plan, this will remain an immutable fact.

Yet part of the journey’s learning has been an advancing sense of detachment with the specific dollar volatility of the portfolio over any short period. This is perhaps as much due to gradual exposure, and a process of acclimatisation, as any particular philosophical achievement or feat of stoicism.

A short history of change in portfolio investment fees

The fees on the earliest investment in what would become the financial independence portfolio were punishingly high, in retrospect.

My first managed fund investments with Colonial First State had management expense ratios of around 2 per cent or more. This effectively resulted in fees consuming a high proportion of long-term expected equity returns.

From March 2001, directing regular investing through Vanguard retail funds provided some relief to this situation.

These funds then operated a tiered fee structure of charging 0.9 per cent for the first invested increment of $50,000. Progressively lower fees applied to amounts over this initial threshold. Heavy investment in these funds increasingly meant that the average investment fee paid fell closer to around 0.3 per cent, from 2001 to 2017. Vanguard has since revised its retail fund fee arrangements to a flatter and more competitive structure.

From 2017, new investments increasingly were directed into exchange traded funds, with Betashare’s A200 and Vanguard’s VAS each offering the opportunity to substantially reduce fees. A200 has an expense ratio of 0.07 per cent, and VAS fees were recently lowered to 0.1 per cent.

As a result of this, the average expense ratio for the entire traditional investment portfolio has fallen to around 0.22 per cent. This falls to around 0.17 per cent if Bitcoin holdings are included.

The growth in the size of the portfolio, however, makes this relatively low expense fee a significant expense item, when taken as a whole.

At the current portfolio level, annual fees of around $4,500 are effectively deducted from investment returns, the equivalent of about $86 per week. Had the earliest managed fund fees I encountered continued to apply, it is worth noting that this would have led to annual fees of northwards of $40,000 per annum.

Observations on a history of portfolio change

As the journey has progressed, and the force of compounding has started to tell upon the portfolio, I have found that the nature of the expected and experienced journey has diverged. The history of portfolio change is in some senses a history of different, and in some cases unanticipated, questions and answers arising.

Once, the challenge was how distant and – thanks to inflation – seemingly ever receding the portfolio goal seemed. At present, at least, the portfolio changes have meant that the challenge is that the intended goal fells pressing. This poses insistent questions such as: what does the achievement of the goal mean in practical terms, and what comes next?

As time goes on, with the signal exception of Bitcoin, I find myself less interested than previously in exploring each new individual investment product or service. Taken as a whole, once a low cost diversified passive investment vehicle with an appropriate asset allocation is selected, there is little to be gained through any one new offering.

This has seen the portfolio re-concentrate around several low cost ETFs and remaining Vanguard funds. These now make up nearly 75 per cent of the portfolio by value and drive the overall equity performance of the portfolio.

The growth of distributions with the size of the portfolio has been the most tangible regular measure of progress to assert itself each quarter. This has gradually brought about changes in how ongoing reinvestment occurs, and a need to provision for material tax liabilities. So far, these changes have been gradual enough in nature to allow a trialling of different approaches, and adjustment.

The portfolio is likely to continue to change, with an emphasis on simplicity of management, and tax efficient simplification of the number of holdings over the next 3 to 5 years.

Following this period, I anticipate a largely ETF centred portfolio will be maintained. A key focus will be maximising geographic, industry, and currency-based diversification which is highly liquid, while maximising real (after tax and inflation) returns.

In the current highly uncertain geopolitical and global asset and investment environment, this has never seemed a more challenging task. After all, tomorrow is a river which has not yet been stepped in.