From Gerd Kommer and Alexander Weis
Since many investment funds - be they actively managed funds or ETFs, stock funds or bond funds - have both a distributing and an accumulating variant, an investor is fundamentally faced with the question of which fund variant he should choose. In the case of distributing funds (hereinafter referred to as “A funds” for simplicity), the interest or dividend payments from the securities in the fund are distributed to the investor after taxes have been deducted. The value of a fund share is reduced by the distributions.
In the case of accumulating funds (hereinafter “T-funds”), the fund also receives this current income from the securities contained therein (stocks, bonds), but does not distribute them directly to the investors, but rather reinvests (accumulates) them in the fund assets free of charge for the investor. As a result, the fund share value increases over time in the amount of this current income minus taxes.
In order to adequately answer the question of which type of distribution from a fund - distribution or accumulation - is preferable from your own personal point of view, you have to distinguish between three different investor constellations, because the answer "which is better for me?" ultimately depends on the particular constellation in which the investor finds himself.
However, before we go into the three constellations, we first have to briefly break down three basic issues.
(a) When an investor withdraws cash from a fund, a portfolio of funds or individual securities, this is economically a “withdrawal”. Whether this withdrawal occurs through the receipt of a dividend or interest payment or through the sale of a fund share or a security (share or bond) is initially economically irrelevant. In all cases, it is the investor's money that flows out of the portfolio or portfolio. Dividends as a source of withdrawals have no advantage over withdrawals via share sales, subject to possible differences in taxes (which we discuss below) and transaction costs (costs of buying and selling). (We will also go into the transaction cost issue in more detail.)
To avoid misunderstandings: All statements about taxes in this blog post assume that the investor is subject to unlimited taxation in Germany, that the fund shares are held as private assets and that the custodian is located in Germany. However, most of the statements we make would also apply without these restrictions.
In one of our next blog posts we will address the “dividend fallacy” that has been widely studied in the academic literature (see here), the false belief that high dividend yield stocks produce higher total returns than equivalent low dividend yield stocks because of their high dividend yield. Two further errors in thinking within the Dividend Fallacy are: For a significant number of private investors, a total return of - let's say 10% - feels more valuable if it consists of 7% price gains and 3% dividends than a total return of "only" 10% price gains. Relatedly, many private investors perceive a withdrawal via dividend as “somehow” less of a reduction in their portfolio assets than a withdrawal of the same amount via the sale of shares. These different perceptions are irrational because they cannot be justified in purely economic terms.
(b) On the tax effect of withdrawals: In general, the tax effect of a cash withdrawal via dividend and a cash withdrawal via share sale in a fund portfolio at investor level is identical or almost identical in the vast majority of normal constellations and if the calculation is methodically correct. This is because the legislature does not want “tax arbitrage opportunities” to exist here. For example, the tax advantage of zero coupon bonds (bonds with no interest but with a higher return) that existed many years ago has been repealed. The fact that there is no significant difference in tax terms between distributing and accumulating funds, between a withdrawal in the form of a distribution or a sale of shares, is trivial, but it must still be stated explicitly, because it is repeatedly shown that some private investors suspect that there are no differences here. If anything, investments in which a high proportion of the total return comes from capital gains rather than current income have a tax advantage, but only if the investor practices strict buy-and-hold over many years. We describe this tax advantage here. (However, this aspect plays no role when weighing up between A funds and T funds.)
(c) By definition, private investors are in one of three possible phases with regard to their portfolio (deposits): the savings phase (asset accumulation phase), the dissaving phase (asset use or consumption) or a “neutral” phase. The savings phase is defined in such a way that over a sufficiently long period of time, more money is paid into the portfolio than is withdrawn (net addition). In the dissaving phase, however, more funds are withdrawn from the portfolio over a sufficiently long period of time than are added (net withdrawal). This net withdrawal by the investor can be smaller or larger than the current income (dividends) or the current total return (dividends and price gains). How high the expected withdrawals will be relative to any expected distributions for A funds is an important technical aspect that is often overlooked in this context. We will return to him below. In the neutral phase, there are no net additions or withdrawals over a longer period of time. Empirically speaking, this phase is likely to be the shortest of the three phases over the entire “life of a private investor”.
Let us now come to the three investor constellations mentioned at the beginning, which largely determine whether a distributing or accumulating fund is preferable for a given investor.
Constellation 1: The investor is in the savings phase (asset accumulation phase)
In this constellation it will almost always make more sense to opt for T-funds. There are four reasons for this, which partially overlap: (a) lower transaction costs because there are no costs for reinvestment, (b) less work required for reinvestment, (c) advantageous self-discipline and (d) obtaining a small return advantage.
While (a) and (b) should be self-explanatory, let's go into points (c) to (d) and go a little further. With (c) - self-discipline - the temptation to waste a distribution for consumption only exists, by definition, with A funds (distributing funds). Some investors may be immune to such temptations, but for them this argument in favor of T-funds is irrelevant, but probably not all of them.
As far as the (small) return advantage (d) is concerned, you have to keep the following in mind: Using an A fund often results in the distributions “lying around” in the zero-yield clearing account (custody account) for a more or less long time until they are reinvested again. Depending on the investor's level of attention, this means lost income (in economist's jargon, "opportunity costs"), because within a T-fund this money would have achieved a higher return on average. The effect may be small, but probably not negligible over a very long period of time.
Constellation 2: The investor is in the dissaving phase (asset utilization phase)
In this constellation, the recommendations for action are somewhat less clear than in constellation 1. Three cases can be distinguished: (a) the investor's annual liquidity requirement is approximately the same as the expected distributions (from all funds combined), (b) it is higher or (c) it is lower.
Since case (a) rarely or never occurs in the long term, we consider it to be unrealistic and therefore no longer relevant for the purposes of our analysis.
For case (b) – the expected annual portfolio withdrawals are always or most of the time higher than the expected annual distributions – T-funds should be slightly preferable. In this case, by definition, additional fund shares must be sold because the distributions alone are not enough. The conceivable advantages of A funds (e.g. convenience and lower transaction costs) largely evaporate. It should be noted that the majority of the transaction costs for share sales in normal private customer business are fixed at most custodian banks, so they depend little on the level of the purchase volume. Small buy or sell orders cost almost as much as large orders.
Now to the last case (c) – the required liquidity (portfolio withdrawal) is lower than the distributions. The argument in favor of A funds could be that there is no work involved in selling fund shares. That's true, but by definition there is work involved in reinvesting the unused distributions that wouldn't exist with a T-fund. In this respect, the argument about the amount of work overall is unlikely to be decisive here. For analogous reasons, the transaction cost argument will not speak for A-funds here, or only to a negligible extent, because although there are no costs for selling shares, there are costs for buying new shares. And as mentioned, most of these costs are fixed and do not depend on the size of the fund position traded.
Constellation 3: The investor is in the neutral phase (net no withdrawals and no additions)
Given what has already been said, it is actually clear for this constellation that T-funds are preferable.
Further points of view
What should you do if only one of the two options for using income exists for a fund in question? As the previous statements have indicated, in most cases the advantages and disadvantages of A funds versus T funds are not economically significant enough to mean that one should forego an otherwise preferred fund simply because it is only available in one of the two forms. This statement assumes that the investor acts strictly rationally. If this condition is met, the simple rule applies: If a fund that is clearly preferred for other reasons does not exist in the preferred income use variant, you should probably still use this fund.
What role does rebalancing play in choosing between A funds and T funds? (For our blog post about rebalancing see here.) Rebalancing has significant advantages within the world of passive, forecast-free investing, and a rational investor should practice it in a disciplined and consistent manner. With regard to the choice of the specific rebalancing method, science does not give any specific recommendations. In other words: rebalancing is good, but the specific method is irrelevant from an ex-ante perspective, provided the rebalancing process is strictly mechanical, i.e. rule-based.
An investor in the portfolio building phase should use his current portfolio allocations (savings rates) for rebalancing purposes, because from a decision-making point of view, rebalancing is free in this way, provided the savings rates are large enough in relation to the existing portfolio. The transaction costs for the portfolio additions would have been incurred even without rebalancing. There is no rebalancing relevance for the trade-off between A funds and T funds in this constellation.
An investor in the portfolio dissaving phase (asset use/consumption) should generally use withdrawals from the portfolio for rebalancing purposes in order to avoid transaction costs that would otherwise have been incurred for rebalancing purposes. Given these considerations, T-funds have the advantage that rebalancing can be carried out more easily and in a more targeted manner by selling shares than with A-funds, as these could lead to cost-disadvantageous scenarios. An example: The distributions exceed the investor's cash needs, but flow from the wrong portfolio component for rebalancing purposes. Now the investor must, on the one hand, reinvest part of the distributions and, on the other hand, sell shares elsewhere at the same time. Compared to a pure T-fund situation, this probably leads to unnecessarily high transaction costs, more complexity and more work.
Finally, for the sake of completeness, we would like to address a pseudo-disadvantage of T-funds: With T-funds, unlike with A-funds, the investor must supply the liquidity that is necessary to settle the above-mentioned advance tax flat rate (i.e. the tax payment) “from outside”, because in this case the liquidity for the tax payment cannot be deducted by the custodian bank from the distributions that take place once or twice a year (if the cumulative tax liability exceeds the amount of the investor-specific exemption order with the relevant custodian). However, this is a purely psychological disadvantage that should not play a role for a rational investor. Whether a tax payment is made from the fund assets or from the investor's other assets is a left-pocket-right-pocket effect and is economically meaningless for purely rational investors.
Conclusion
This blog post wanted to show when distributing funds and when accumulating funds should be preferred. The specific answer to the investor question “distributing or accumulating funds – which are preferable?” depends primarily on whether an investor is in the savings or consumption phase and how high any withdrawals expected in the foreseeable future will be relative to the expected distributions of an eligible distributing fund.
In general, our scenario analysis has shown that (a) the cost and tax differences for investors between distributing and accumulating funds are probably often overestimated and that (b) most rational investors do slightly better with accumulating funds in many relevant constellations and in some significantly better.
The emphasis is on “rationally acting investors”. From an emotional perspective, distributing funds have the perceived advantages for “irrational” investors that they require no or fewer share sales, they can enjoy dividend payments to the clearing account more than share price increases of the same amount and that they have to obtain no or less tax liquidity from outside the portfolio.