3

I'm currently invested in an open-ended fund. As I understand it, neither I nor my broker see its dividend. We just wake up on the payment day and see that the unit price of the fund has moved more than we'd expect from the movements of the market, and that's that. Our unit number doesn't increase and we don't get any special notices. On dividend day, we don't get told anything more than what we do on any other market day. The unit price gets reported, and that's it. I believe this to be one such fund, as presented by a fairly typical broker.

Assuming that what I've said is accurate, why does the ex-dividend date matter to anyone? It seems that the only change that happens on the dividend date is that the unit price moves, so as far as I can tell, there shouldn't be any difference between someone who buys at price x a week before the ex-dividend date and someone else who also buys at price x a week after. Because they both hold units on the dividend date, it appears obvious to me that they will both benefit from the increase in unit price on that day, independent of when they bought.

Answers so far have pointed out that this is in fact an OEIC (not an ETF as the question originally had it) and that the dividends are constantly reinvested, making the dividend and ex-dividend date meaningless. However, I have three objections to this:

  1. Doesn't this make taxing your dividends - a legally required step - impossible?
  2. Official documents appear to contradict the idea of constant reinvestment. For example, as I read it, page 178 of this report says that the dividends paid on the dividend date for the fund linked above was 20.828558p per share (net) and page 194 informs us that it has been retained in the fund. It also provides these figures for their other funds, furthering the idea that this is not unusual.
  3. If these dates are meaningless, why does anyone bother with equalisation? This is such a big point that it might even be enough to single-handedly answer this question, so I distrust any explanations that overlook it.
OJFord
  • 123
  • 4
J. Mini
  • 607
  • 1
  • 6
  • 16
  • The fund you link to is not an ETF. Note: "Dealing Frequency: Daily" on the page you link. – David Sep 11 '20 at 09:42
  • @David If that's the case, should I be using a different tag? – J. Mini Sep 11 '20 at 14:44
  • @J.Mini, here's some info that may help explain the difference, consider editing the tags and question text. https://www.barclays.co.uk/smart-investor/investments-explained/funds-etfs-and-investment-trusts/whats-the-difference-between-an-etf-and-a-tracker-fund/ – C8H10N4O2 Sep 11 '20 at 15:21
  • @C8H10N4O2 Thanks, but I see no OEIC tag. Should I be looking elsewhere? – J. Mini Sep 11 '20 at 16:16
  • Investopedia says closest comparison in the USA is an open-ended mutual fund (I think OE ones are rare though) - so I swapped ETF tag for mutual funds; someone with more rep could add a new tag, such as 'open-ended'. – OJFord Sep 11 '20 at 18:03

5 Answers5

10

The fund you linked is described as "accumulation units" and has a counterpart described as "income units". Most ETFs (in the US in particular) behave like income units. Accumulation units, which directly reinvest dividends as they are received from stock holdings and do not pay them out in cash, may be common in the UK but are likely obscure elsewhere.

I think you are correct that the ex-dividend date is not meaningful in this case. It may be listed for that fund by default based on the corresponding income units. See discussion here and here.

It seems that the only change that happens on the dividend date is that the unit price moves

It would not move in any outsized way; rather, the price moves every day by the moves in market prices of holdings, plus any dividends paid by those holdings, which would be a very small amount because only a small percentage of stocks pay dividends on any given day. Moreover, those dividends are offset on average by drops in the corresponding stock prices. In effect, the accumulation ETF price tracks a total return index.

It would not be feasible for the dividends to somehow accumulate "outside" until a fixed date (ex-dividend) and only then appear as an extra jump in the ETF price, because everyone would buy the ETF just before and sell it after. No risk-free "dividend capture" arbitrage is possible whether with accumulation units or income units.

nanoman
  • 29,638
  • 3
  • 69
  • 89
  • ETFs that reinvest directly are common in countries that have high taxes on dividend income. This makes it easier to accumulate cumulative gains before eventually paying taxes when selling. – jpa Sep 11 '20 at 13:39
  • @jpa There seems to be some confusion on how the tax man sees accumulation units. – J. Mini Sep 11 '20 at 15:07
  • @J.Mini Interesting, that then also varies by locale. In Finland accumulation units have always been taxed only at the time of sale/withdrawal, while this year the government introduced tax-sheltered investment accounts that allow the same for common stocks and income units. – jpa Sep 11 '20 at 15:48
  • I find myself disagreeing. See the edit to my question. The fund's own documents list it as having a dividend payable on the dividend date. – J. Mini Sep 11 '20 at 16:16
  • @J.Mini I think that can be explained as an accounting formality. Note that on page 194, "distributions" and "retained distributions" cancel. It is essentially a label for an amount that the fund is paying to itself. There appears to be no practical effect to the shareholder on ex-dividend day. – nanoman Sep 11 '20 at 16:45
  • Looking at it that way, the dividend date itself is an accounting formality. Surely that's odd? – J. Mini Sep 11 '20 at 16:51
2

First of all, that's not an ETF (Exchange-Traded Fund) - it's an OEIC (Open-Ended Investment Company). (It's price reflects the underlying holdings, 'open-ended' means it can create and sell you 1000 units without moving the price, assuming the underlying prices don't move.)

You receive dividends from it, the OEIC, not (directly) the underlying - or you would have thousands of ex-div dates, not one.

However, you've linked to the 'Class C - Accumulation' variant, which means that income from the underlying will be reinvested, not distributed as income ('Other unit types available: Class C - Income') - it stays within the fund.

Ex-dividend date thus matters to 'Income' units in the same way as for any company, fund or not. 'Accumulation' units just remove the choice of what you do with your dividend: you buy more shares. Though since the only shares are 'inside' the fund, you're not issued more units, but the units you have represent a claim on more shares of the underlying.

Regarding your updated questions:

  1. You don't actually receive a dividend (with Acc. units) so you have nothing to report even if it exceeded the allowance.
  2. That is interesting/a slightly odd way to word it, but I believe it's just saying that is the value that would be paid out, and the value that will be reinvested. Note that it describes a 'distribution'; not a 'dividend'.
  3. The dates aren't meaningless, because something's really happening: 'Income' units don't sell shares to give you income, it comes from cash; 'Accumulation' units instead use that cash to buy more shares.
OJFord
  • 123
  • 4
  • This appears to repeat the question. If the income is reinvested straight back in to the OEIC, how is there any difference between buying before the ex-dividend date and buying after? Regardless, your points have made me look further in to this issue and have caused some edits to my question that you may find relevant. – J. Mini Sep 11 '20 at 16:20
  • @J.Mini You might expect a bit of extra movement around that period, since the fund becomes less 'bag of European equities plus cash' and more 'bag of European equities', but sure, not much. It makes more difference to 'Income' units, in which case it can be considered no different to the dividend period of any listed (fund or not) company; 'Accumulation' just removes the choice of what you do with your dividend: you buy more shares. (Though note the only shares are 'inside' the fund, you're not issued more units, but the units you have represent a claim on more shares of the underlying.) – OJFord Sep 11 '20 at 17:43
  • I suspect that I'll be accepting this answer once I'm done checking some details. However, your point #3 doesn't appear to explain the importance of the ex-dividend date. It appears to only explain the dividend date. I think that equalisation is the important missed point. – J. Mini Sep 11 '20 at 18:02
  • The fund's cash balance is included in the valuation; so when you buy ex dividend you receive that 'overpayment' back. HL shows this as a 'corporate action' in the 'transactions' table on the 'stock movements' page. (URLs like https://online.hl.co.uk/my-accounts/security_movements/sedol/[ID] from clicking on the name of the fund in 'account summary'.) – OJFord Sep 11 '20 at 18:42
  • I fully agree about the overpayment, see my recent answer. – J. Mini Sep 11 '20 at 18:53
1

Where did you get this impression?
Of course you get a dividend.

Typically four times a year, but some of them only annually, or twice a year, or monthly. For example VTI (https://investor.vanguard.com/etf/profile/distributions/vti):

Type    Distrib     Record     Ex-Div    Payable 
 Div    $0.69990   06/26/20   06/25/20   06/30/20 
 Div    $0.61360   03/27/20   03/26/20   03/31/20 
 Div    $0.88550   12/26/19   12/24/19   12/30/19 
 Div    $0.70000   09/17/19   09/16/19   09/19/19 
 Div    $0.54720   06/18/19   06/17/19   06/20/19 
 Div    $0.77200   03/26/19   03/25/19   03/28/19
Bob Baerker
  • 76,304
  • 15
  • 99
  • 175
Aganju
  • 37,683
  • 7
  • 57
  • 119
  • 1
    Your formatting was a problem. If my attempt to fix it is unsatisfactory, edit to your liking. – Bob Baerker Sep 11 '20 at 00:32
  • 2
    This does not really address the question, which is specific to UK "accumulation units" that treat dividends differently from other ETFs. – nanoman Sep 11 '20 at 02:30
  • Then - maybe the OP should add a tag 'UK'? Looks to me like he talks about ETFs and dividends. There is no mention of UK specific other things. – Aganju Sep 11 '20 at 03:22
  • 3
    Done. OP's link is to a UK accumulation ETF, and the peculiarity of the question (what probably seems to you like confusion) reflects how such ETFs work. – nanoman Sep 11 '20 at 03:34
  • 1
    In most countries, ETFs are legally required to pay out dividends they collect from the shares in them. Otherwise, you could delay paying taxes indefinitely. It is quite surprising that the UK allows that. – Aganju Sep 11 '20 at 03:52
  • 1
    @Aganju I cannot speak for UK, but in Germany, "ACC" ETFs are taxed yearly if they have gained in value. Then, a certain "base interest" is assumed. This already paid tax is then subtracted if you sell the ETF one day. – glglgl Sep 11 '20 at 08:12
  • This is not an ETF. (And that's not just being pedantic about terminology - it's not Exchange-Traded.) – OJFord Sep 11 '20 at 14:20
1

I have no clue how dividends are handled in the UK. Here's how they are handled in the US. See my answer here.

If dividends are reinvested, they are of no consequence other than taxation if received in a non sheltered account and possibly a minor discrepancy in reinvested shares acquired because the Payable date is after the Ex=Dividend date (share price will likely be marginally higher or lower).

Bob Baerker
  • 76,304
  • 15
  • 99
  • 175
1

It's bad form to answer my own question, but the comments have reveals that the question was flawed. First of all, OJFord is right to point out that this is an OEIC and not an ETF. Furthermore, the question assumed that the dividend date was when the dividends accumulated from the fund's underlying stocks were reinvested. However, if you compare any accumulation class fund's percentage change in price (and not total return!) with its income class equivalent, you will see that this is usually false. For example, with the accumulation class in red and the income class in blue, this graph is what we get from the question's linked fund, showing that the divergence was only triggered by the ex-dividend date (mid May, with the dividend in mid July). This also implies that the dividends from the accumulation class were reinvested on the ex-dividend date, destroying the premise of the question.

So why does this confusion exist? This takes some background work. Ignoring the ex-dividend date for a moment, here is my understanding of how the dividend payments work for accumulation units in an OEIC:

  • As stated in the question, for an accumulation class fund neither you nor your broker get told what the value of the dividend was. There is a date when it's supposed to become relevant, the dividend date, but this is misleading because you won't see the dividend and it may have been reinvested at any prior point on or after the ex-dividend date. Ultimately, for accumulation type funds, the dividend date is little more than a tax/accounting detail.
  • If you want to find out what the dividend's value really was, you need to find the official documents from the fund. As with the example linked in the question, these confirm that a dividend exists and is reinvested back in to the fund. They also give its value. Incidentally, it's unclear to me if the tax man cares about this, so beware!

With this key fact - that the dividend is never seen leaving the fund - established, we can now ask the question that we were trying to answer: If the only thing that changes due to the fund's dividend is the unit price of the fund, why does the ex-dividend date matter? After all, I'll benefit from the change in unit price regardless of if I buy before of after the ex-dividend date.

Our answer to this is equalisation. That link explains it better than I can, but here is how it applies to this question:

  • Over time, the fund will accumulate cash due to the dividends that are paid to it from the shares that it owns. These are not reinvested until at least the ex-dividend date. However, the fact that the fund has this cash increases the value of the fund. Of course, some of this cash will be used to cover costs like staff expenses, but the important part is that...
  • On the ex-dividend date, a great deal of this cash is set aside for the purposes of dividend payment (or in our case of accumulation units, dividend reinvestment). As this cash is now committed somewhere, unless it is reinvested back in to the fund immediately (as it was in the question's linked example), the value of the fund goes down.
  • However, the fund will probably still be getting dividends paid to it from its underlying shares, so even before the dividend date, the value of the fund will start going back up. This continues until the next ex-dividend date.
  • This means that if you buy in to the fund between two ex-dividend dates, you are paying extra because of the cash that the fund has got sitting around.

Technically speaking, this has already answered the question. The ex-dividend date matters because it has a clear effect on the price of the fund. However, there is a final loose end that we should clear up.

As this question points out, everyone benefits equally from the change in unit price when the dividends are reinvested. However, the funds recognise your overpayment and equalisation is the fancy name for how they compensate for this. On or very near the date of reinvestment, the amount that you are recorded as having paid for your investment will be revised down in proportion to how much you overpaid (i.e. your proximity to the upcoming ex-dividend date). In my experience, your unit number and price will not change when this happens and your account won't be credited either. It's always been a case of the fund saying "we recognise that you overpaid by x amount, so we've reinvested it back in to the fund. As you're invested with us, everybody wins". And that's the end of the story!

In summary:

  • Yes, everyone benefits equally (per share) from the dividend regardless of when they buy, as long as they buy before it is reinvested.
  • No, this does not mean that the ex-dividend date doesn't matter. The unit price of the fund will typically drop notably on it and you will overpay (albeit with some unusual compensation) if you buy in after it.
J. Mini
  • 607
  • 1
  • 6
  • 16
  • "The ex-dividend date matters because it has a clear effect on the price of the fund." This is true for income units (which that equalisation page seems to focus on), but not for accumulation units. On the fund page linked in your question, if you click to the corresponding income units and bring up a 6-month percentage chart for both (under "Charts & Performance"), you will see that the income and accumulation units have identical total return, and the price chart of the accumulation units is likewise the same, but the price chart of the income units differs due to an ex-dividend in May 2020. – nanoman Sep 11 '20 at 19:57
  • The link's explanation is odd. It suggests that the NAV increases because the fund is receiving dividends from the components so it increases in price yet they acknowledge that the fund's share price decreases (see graph) when the dividend is paid out (setting aside cash to be paid out). When a fund's components go ex-div, the NAV drops and this is ignored in the explanation If share price is not changing due to market forces, NAV will be lower by the amount of the distribution. IOW, Pre = Post + Div with no difference in value if reinvested. If dividend withdrawn, value must be lower. – Bob Baerker Sep 11 '20 at 20:00
  • @nanoman 6 months? Aren't these dividends annual? Why 6 months? – J. Mini Sep 11 '20 at 20:10
  • @BobBaerker The graph on the equalisation page appears to assume that, over time, the portfolio is appreciating in price at a similar rate as the dividends are reducing it. On any given day that is not ex-div for the fund, most holdings go up slightly on average while the few that are themselves ex-div go down on average, resulting in a small positive return after adding in the dividends received (or at least accrued) by the fund. The sawtooth graph would be seen in purest form in something like a short-term bond fund (with interest instead of dividends). – nanoman Sep 11 '20 at 20:12
  • @J.Mini The last ex-div was in May, so a 6-month chart suffices at the moment to show it. – nanoman Sep 11 '20 at 20:14
  • @nanoman It's hard to compare the 6-month price charts because they've got different scales on their y-axis, but I think that I can see it. Also, why can't I see any difference in the 3-month price charts? Shouldn't the dividends make a difference? – J. Mini Sep 11 '20 at 20:46
  • @nanoman - If they are including an assumption of fund appreciation over time in an explanation of how they handle distributions then it's not odd, it's a bad explanation. The two components should be segregated because combining them prevents a reader from understanding exactly how dividends and distributions are handled. – Bob Baerker Sep 11 '20 at 20:49
  • @BobBaerker Are you sure that they acknowledge that the price decreases when the dividend is paid out? To me, it looks like they're talking about the ex-dividend day that their graph just happens to have one day after the dividend payment. – J. Mini Sep 11 '20 at 20:58
  • @nanoman The income and accumulation units definitely don't have the same return. You can put them on a graph and see this easily (red is accu, blue is income). https://i.imgur.com/S1txHfR.png – J. Mini Sep 11 '20 at 21:11
  • @J.Mini Good graph -- that is exactly my point. I think once you understand this, you'll be close to a complete answer! For "chart type" you have chosen "price" rather than "total return". You are precisely seeing where the price of income units dropped compared to accumulation units on that day in May, but thenceforth followed the same course. The total return charts are, however, identical (and the same as the accumulation price chart). 3-month charts don't currently show the difference because they don't span an ex-div date. (This date is what matters, not the payment date in July.) – nanoman Sep 11 '20 at 21:20
  • @nanoman But on the payment date, shouldn't the price of the accumulation fund move up more than the income? – J. Mini Sep 11 '20 at 21:24
  • @J.Mini Regarding equalisation page: If you're interpreting that there is a dividend payment on date C, I disagree. Date C is just shown as the last possible day to buy the fund and qualify for the dividend. The actual payment date is not shown and not important to the explanation. Regarding HSBC fund payment date: For accumulation units, the dividends are already part of the fund's value. That the total returns are identical indicates that the dividends are reinvested on the fund's ex-div date. Nothing special happens on the payment date except income unit shareholders receive their cash. – nanoman Sep 11 '20 at 21:36
  • @nanoman Nothing special happens on the payment date except income unit shareholders receive their cash. Sorry, I just find that plainly absurd. It's what the charts say, but I can't believe it. Do you have any sources indicating that this is normal practice? As a side note, this discussion should probably be put in to an answer somewhere rather than left as a huge discussion in the comments. – J. Mini Sep 11 '20 at 21:50
  • @nanoman So, all in all, the actual dividend date is just something to be noted for accounting/tax purposes? – J. Mini Sep 11 '20 at 22:37
  • @J.Mini See here and here for (US-focused) explanations of dividend dates. The payment date reflects the time needed to "cut the cheques" after the list of dividend-eligible shareholders is determined. As I noted, accumulation units appear to do their reinvesting on the ex-dividend date. So the dividends are removed from the fund's cash balance (just as for income units) and immediately contributed to the stock portfolio (instead of cutting cheques as for income units). – nanoman Sep 11 '20 at 22:46
  • @nanoman I think we're talking past each other. What I mean to say is that it appears to me that outside of noting things for accounting/tax purposes, it appears to me that the dividend date is meaningless for the accumulation units. – J. Mini Sep 11 '20 at 23:15
  • I own a $100 stock that goes ex-div for $1 tomorrow. In the absence of price change due to buying or selling, tomorrow I own a $99 stock and owed $1 on Pay Date. $99 + $1 = $100 regardless of whether I take the cash or reinvest. Suppose fund owns 50 different stocks doing this every month. Every stock is the same on the ex-date. In the hypothetical situation where no stock changes in price other than ex-div reduction, the fund would experience the same. Asset value would drop slightly on every stock's ex-date but NAV would remain the same since the ex-div loss is recovered on the Pay Date. – Bob Baerker Sep 11 '20 at 23:18
  • @J.Mini I agree -- I was just adding (perhaps unnecessary) elaboration that might clarify things because you found the situation "plainly absurd". – nanoman Sep 11 '20 at 23:41
  • @nanoman Attempted to edit answer to reflect this discussion. Will read back over it once I've slept on this. – J. Mini Sep 12 '20 at 00:13
  • "It's bad form to answer my own question," No, it isn't. – glglgl Nov 07 '20 at 20:37