A fund's return is the weighted return of its holdings less what it charges. Knowing that changes what you are actually forecasting when you forecast a fund.
11 min read · Level 2
The most common reason given for buying a fund is that it removes the need to have a view, and that is not what happens. It relocates the view. Somebody still decided that these five hundred companies are the eligible universe, that market capitalisation is the right way to weight them, that the list is reconstituted in March and September, and that a company is dropped when it falls below a threshold nobody at the fund chose either. Those decisions are a set of opinions written down as rules, and buying the fund adopts every one of them. What the holder has genuinely escaped is the need to form a new opinion each week, which is a real and large benefit - improvised views formed under pressure are worse than published rules followed without emotion, and the evidence on that is not close. But the honest version of the claim is not that the fund has no view. It is that its view is stable, disclosed, cheap and not yours, and that on balance those four properties beat the alternative for most people most of the time. Anyone who buys a fund believing they have opted out of judgement should read the index methodology once. It is usually about thirty pages, and every page is a judgement.
A fund's return is not produced by the fund. It is the weighted average of what its holdings did, reduced by what the fund charged to hold them. Write it out once on a portfolio small enough to see.
A five-holding fund, at the start of the year:
Holding weight year's return contribution H1 40% +12% +4.80% H2 25% -4% -1.00% H3 20% +7% +1.40% H4 10% +25% +2.50% H5 5% -30% -1.50% ------------------------------------------------ Gross portfolio return +6.20%
Now charge an expense ratio of 0.30%, levied on the value at the end of the year:
1.0620 x (1 - 0.0030) - 1 = 5.8814%
Two things are worth pausing on. The first is that 6.20% is the entire story of the year, and it contains a holding that fell 30% and a holding that rose 25%, neither of which is visible in the headline. A fund's published return is a number that has already hidden everything interesting about its constituents.
The second is what happens if you read the same five holdings equally rather than by weight: (12 - 4 + 7 + 25 - 30) / 5 = 2.00%. The same five companies, the same year, and a 4.20 percentage point difference between the two readings. Weighting is not bookkeeping. It is most of the answer.
There are exactly three things the wrapper can do to the number that comes out.
It subtracts its costs. That is the expense ratio, plus dealing costs incurred inside the fund, which the expense ratio does not include.
It changes when tax falls due, sometimes considerably, which lesson five takes apart.
And in a fund that lends its securities to other market participants for a fee, it can add a small income the holdings themselves did not produce. That income is real, it is usually shared with the fund's manager, and it comes with exposure to whoever borrowed the securities.
That is the complete list. There is no fourth item in which skill at packaging creates return.
Because it settles what you are actually forecasting when you form a view on a fund, and the answer is uncomfortable: you are forecasting a basket of assets you did not choose and have probably not examined, at weights set by a rule you did not write.
It also isolates the one variable in the whole chain that is known in advance. Suppose the underlying basket returns 7.00% in a year, with the fee charged on the year-end value.
Fund A, expense ratio 0.05%: 1.07 x (1 - 0.0005) - 1 = 6.9465% Fund B, expense ratio 0.85%: 1.07 x (1 - 0.0085) - 1 = 6.0905%
The fee difference is 0.80 percentage points. The return difference is 0.8560 percentage points, because the fee is levied on a value that has already grown. The gap is slightly larger than the headline difference in fees, not slightly smaller, and it widens with the return.
Then there is the route the return takes to reach you, which is a property of the share class rather than of the portfolio.
A distributing class pays income out. A dividend received by the fund goes to your account as cash, and the unit price drops by the amount paid.
An accumulating class retains it. The same dividend stays inside the fund, is reflected in a higher unit price, and compounds without you doing anything.
The total return is identical. What differs is whether the compounding happens automatically or requires you to act - and the arithmetic of that difference is the same arithmetic a bondholder faces with coupons. Assume the holdings pay 2% of the year's starting value as dividends and appreciate 5%, so the total return is 7%. Over ten years, per 1.00 committed:
Accumulating, everything reinvested Ending value 1.967151 Annualised 7.0000%
Distributing, income taken and spent Capital 1.628895 Income collected 0.251558 (0.02 x [(1.05^10 - 1) / 0.05]) Total 1.880452 Annualised 6.5188%
A gap of 48 basis points a year, on identical holdings, with nothing having gone wrong. It is not a defect in the distributing class. It is what spending the income costs, and spending the income is frequently the entire reason somebody holds the fund. The error is only ever in the accounting afterwards.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
The consequence that catches people out is that the two classes do not look the same on a chart, and the chart is what most comparisons are made from.
Take one year on Fund A's units. Suppose net asset value starts at 50.00, the holdings return 7.00% of which 2.00 percentage points arrives as dividends, and the fund charges 0.05%.
Accumulating class 50.00 x 1.07 x (1 - 0.0005) = 53.4733 Holder's position: 53.4733 of units, no cash
Distributing class Same portfolio, same fee. Income of 2% of 50.00 = 1.00 per unit is paid out. Unit price: 53.4733 - 1.00 = 52.4733 Holder's position: 52.4733 of units, plus 1.00 of cash = 53.4733
Identical outcome. But a price chart of the two classes shows one ending at 53.4733 and the other at 52.4733 - the accumulating class looks 1.9057% better over a single year, and the gap compounds visually over a decade until the distributing class appears to be a materially worse fund.
It is not. The difference is money that was handed to the holder. Any comparison of two funds must be made on total return, which adds distributions back, and a surprising number of freely available charts do not.
There is a sharper version of the same confusion. A fund can distribute more in a year than its holdings earned, by paying out capital alongside income. The cash arriving looks like a generous yield; the unit price falls by more than the market alone explains; and the holder, watching a healthy income and a declining price, concludes the manager is losing money. What is actually happening is that they are being handed their own capital back and taxed on part of it. The fund documents disclose this, under a heading almost nobody reads, and the distribution yield on the marketing page does not.
Comparing two funds on their price charts when one of them distributes. The distributing one will always look worse, permanently, and the difference is cash that was paid to its holders. Use total return or compare nothing.
Believing a high distribution yield is a high return. A distribution is a transfer from the unit price into your account, not a gain. The only figure that describes what the holding did is total return, and a fund distributing part of its capital can show a rising yield while shrinking.
Attributing the return to the wrapper. When a broad fund has a strong year it is because its holdings did; the wrapper contributed a subtraction. This matters practically, because it means the question "is this a good fund?" cannot be answered without asking "compared with what its holdings did", which is exactly what tracking difference measures in lesson four.
Reading a headline return as a description of the constituents. The five-holding fund above returned 6.20% while containing a holding down 30% and one up 25%. A quiet aggregate can sit on top of a violent set of parts, and the aggregate is the only number most holders ever see.
Letting distributions accumulate as idle cash. If income is not going to be spent, leaving it uninvested produces the 6.5188% outcome above while the holder believes they are getting 7.0000%. Either spend it deliberately or reinvest it deliberately; the default of doing neither is the one nobody chose.
A fund's published return is already net of its expense ratio. Add the ratio back before comparing it with an index, or you are comparing a net number with a gross one and the fund will look worse than it was. For anything more precise than that, compare against the index the fund actually tracks rather than a convenient broad one.
Price return excludes distributions; total return includes them reinvested. For an accumulating class the two figures are nearly the same and for a distributing class they diverge permanently. Whenever two funds look implausibly different over five years, this is the first thing to check and it resolves the discrepancy more often than any other single cause.
Reinvested into the same fund, swept into cash, or spent. All three are legitimate, and all three produce different outcomes on identical holdings. Set the instruction with the broker at the point of purchase, because the default is usually cash and cash is the one option nobody selected on purpose.
A 0.80 percentage point difference in expense ratios produced a 0.8560 point difference in return above, because the charge lands on a value that has already grown. Whenever the basket rises, the direction of the discrepancy is the same - the gap is larger than the fee difference - and over long holds it compounds, which is the subject of fee-drag-over-a-long-hold.
Take a fund you hold, find the returns of its ten largest holdings over a year, weight them, and see how much of the fund's whole result those ten explain. For a fund with 35.8% in its top ten the answer is startling, and after you have done it once you will never again read a fund's return as a statement about several hundred companies.
Whenever you are forming a view on what a fund might do, whenever you are comparing two funds that hold different things, and whenever you are deciding between an accumulating and a distributing class of the same portfolio.
Do not use a fund's past return as evidence about the fund rather than about its holdings, and do not compare funds on distribution yield at all - it is a decision about where money sits, not a measure of what was earned.
Comparison tools that plot unit prices rather than total returns will show a distributing share class trailing its accumulating twin by roughly the income yield, compounding, forever. On a portfolio yielding 2% a year, a decade of that is a visual gap large enough that holders regularly conclude the two are different strategies or that one carries a hidden charge. Both classes hold the same portfolio, and the difference is cash that left the fund and arrived in the holder's account. The reason this misreading is durable is that the chart is free and instant while the total return figure takes a further click, and the chart is not marked as showing price only.
Before concluding one fund has beaten another, confirm both figures are total returns. Half the surprising gaps between similar funds are this and nothing else.
Funds sold on their income sometimes pay out more than their holdings earn, funding the difference from capital. The reported distribution yield is high and keeps being met; the unit price drifts down by more than the market explains; and because the income arrives faithfully every quarter, nothing prompts a review. Holders in this position often describe the fund as performing well on income and poorly on capital, which treats one event as two. The disclosure exists - the annual report states what part of distributions represented a return of capital - and it sits in a section that the marketing material never references.
Check what proportion of a fund's distributions came from income rather than capital. A yield paid out of your own money is not a yield.
Create a free account to read the rest and apply it to your own investing.