<<< This blog post is also available as a YouTube video. >>>
From Gerd Kommer and Alexander Weis
When private investors want to make a stock investment, they often ask themselves timing questions like the following:
- Should I enter the market right now with my entire investment amount?
- Shouldn't I spread my entry into the stock market bit by bit over a longer period of time?
- In principle, couldn't it be better to do nothing at all and wait?
In this blog post we primarily address the first two questions, but also address the last one.
Our question is therefore: (a) Immediate entry via a one-off investment (hereinafter abbreviated as “SPE"), (b) Extended entry as a "phase investment" over 12, 24 or 36 months (hereinafter "GEP“) or (c) do not set a specific entry mode for the time being and wait until further notice (“BAW“).We analyze all three alternatives from eleven different perspectives.
Aspect 1: The statistical expected return
From a scientific perspective, there is no doubt that the SPE approach has a higher expected return looking forward than the GEP approach. This finding is statistically clear; you could also say black and white. What is behind this statement? Given normal valuation conditions, shares have an expected return that is around seven times as high as a “savings book”, i.e. a “risk-free” investment (very short-term government bonds with a high credit rating). Secondly, stocks always have an expected return > 0, regardless of their valuation level. This means that “every day” that you are not invested, you miss out on returns or returns in the sense of the expected value. Thirdly, share prices cannot be predicted reliably enough in the short and medium term (over a period of one day to around five years) to be able to achieve a reliable excess return compared to the market average using price forecasts based on costs and risks. That applies too everyone Time; It doesn't matter whether prices have fallen by 40% in the last six months or have risen by 200% in the last five years. The short and medium-term unpredictability of share prices (Jacobs/Weber 2016) is difficult to mentally accept for many investors, but that does not change the facts. If you combine the always positive return expectation and the unpredictability, the logical conclusion is: SPE must be statistically more profitable than GEP. Historical-empirical data confirm this finding - figures on this below.
Viewpoint 2: Empirical return data
The expected return mentioned above is a statistical concept and that is where the problem for the human psyche begins. We would like to explain this using some figures and look at the global stock market over the last 94.4 years - as far back as the data available to us goes back (January 1926 to the end of May 2020). [1] These 94.4 years correspond to 1,133 months. We imagine two private investors: Anna and Robert. On each of these 1,133 months, Anna invests 100,000 euros (or monetary units in general) on the last working day of the month shortly before the stock market closes according to the SPE approach, while Robert, for reasons of caution, invests his 100,000 euros in the market spread over 36 months (GEP approach). So he invests one thirty-sixth (around 2,780 euros) on the “starting day” when Anna invests 100,000 euros. Exactly one month later, Robert invests the next 2,780 euros and so on until he has invested the last installment 35 months after the starting day. We repeat this experimental set-up for each of the 1,133 months and thus essentially have 1,133 comparison cases for which we can calculate Anna's and Robert's returns over a time window of 36 months each. At the end of the 36 months, by definition, both are fully invested and therefore have the same returns for the following period, which therefore no longer interests us. Since there are no longer any full three-year periods in the last 35 months up to May 2020 (and GEP can therefore no longer be fully implemented), we ignore these final periods in our analysis. This leaves 1,098 comparable cases (1,133 minus 35). Now we evaluate which of the Anna and Robert is ahead more often in the 1,098 complete three-year periods. Result: Anna wins the race 73% of the time. She is ahead more often and the average real final asset value after 36 months is around 126,000 euros compared to only around 113,000 euros for Robert. On average, this is a cumulative additional return of 11.5%. [2] Of course, in the 27% of all cases that go in Robert's favor, there are also a few in which Robert's return advantage (more precisely, his relative loss advantage) is high, but all in all, the overall conclusion remains clear: the SPE approach clearly beats the GEP approach.
Viewpoint 3: Psychological factors
Now, however, a real private investor, let's call her Juliane, doesn't have 1,098 attempts, but only one. Unlike Anna, the law of large numbers will only help her to a limited extent. Statistically and rationally speaking, Juliane would also have to follow the SPE approach. But if her only attempt falls into the 27% pot, she will probably be grieved and regretful for some time - if she is a normal private investor. Depending on your personal nature and the specific return result, it may even hurt you a lot, because as we all know, losses hurt more than gains are good. This also applies, somewhat to a lesser extent, to: relative Losses, i.e. lower positive returns than a subjectively chosen comparison brand. Because this is the case, most private investors who are not strongly rational and data-driven are likely to do so Homo economicus In the interests of their “psychological rationality”, it is better to use the GEP method, i.e. to stretch out their stock market entry mechanically over 12 to 36 months.
Viewpoint 4: Absence of “time diversification”
The statement often heard in the financial industry and the financial media that the longer the time window, the more the return differences for certain observation periods become more similar is not true. In other words: Even with very long observation periods of e.g. B. 20 or 30 years, the specific entry point (and exit point) plays a significant role in the final asset value. The return adjustment in terms of the final asset value (cumulative return) does not take place. Seen this way, there is no “time diversification” and no “time heals all wounds”. In this view, equity risk does not decrease with the length of the investment horizon and the timing of entry is important in terms of returns even over long horizons. However, this does not change the crystal clear statement made earlier that the SPE method is statistically and empirically superior to the GEP method and that an exclusively rational investor would use it everyone would prefer market situations over GEP. (However, time diversification exists in the sense that the “shortfall risk” decreases with the length of the investment period for stocks. Shortfall risk is the risk of not achieving a given minimum return.) One should also add here that in the real world, the vast majority of private investors repeatedly add or withdraw cash from the portfolio over periods of 5+ years. Where this is the case, the impact on returns of the initial entry point, which is the primary focus of this text, actually decreases.
Point of view 5: The exit time is just as important as the entry time
Many private investors struggle with the exact timing of their market entry into stocks because they are afraid of receiving a comparatively poor return or even a loss in the next 12 to 60 months. However, for the total return that is their focus of concern over a given period of time – be it six months or 30 years – it is the timing of the exitpoint in time (or, more generally, the end point of the calculation) is not a bit less important. Nevertheless, the majority of private investors have no or almost no problem with exactly when they sell their stock investments, i.e. with the timing of their exit. This mental asymmetry is one of the endless examples of how irrational we can be in economic matters. In the following point 6 we describe another such example.
Viewpoint 6: Consistency in investment decisions
Private investor Oliver made an investment in an MSCI World ETF a good six years ago in April 2014. The system is now worth 100,000 euros. In these six-plus years, Oliver achieved roughly the historical long-term global stock market return (adjusted for inflation, around 5.5% p.a. before taxes and costs), with which he is satisfied. A few months ago, Oliver inherited another 200,000 euros from his great uncle. The money has just been credited to his account and the inheritance tax has already been paid. Oliver would like to increase his MSCI World investment by the full amount of 200,000 euros, but is afraid of getting caught at a bad time. The further effects of the current Corona crisis on the stock market currently (June 2020) appear unclear and threatening to him. That's why he wants to wait and see - he doesn't want to commit to how long and until what event. This makes Oliver an excellent example of a common form of investor irrationality that is reminiscent of a moderate form of schizophrenia. If Oliver acted logically and consistently, he would either have to immediately withdraw the first 100,000 euros from the market (sell his ETF shares) or immediately invest the new 200,000 euros in its entirety. Oliver does neither one nor the other. Instead, he measures what he sees as the equity risk for the invested 100,000 euros differently than for the 200,000 euros that have not yet been invested. Irrational, inconsistent and at the same time completely normal, because the majority of all private investors probably act this way.
Viewpoint 7: Market crash timing
If the markets have trended noticeably upwards over several years, then the reliability of the tower clock of St. Peter's Church in Zurich is a popular topic in the private investor community Crash timing on. This is the idea that you can achieve above-average returns by spending a certain amount of money waiting for a sharp drop in prices and only then entering the market. Many scientists have grappled with this idea over the last few decades and almost all of them came to the same negative conclusion: crash timing is economically unattractive. We ourselves have in our blog post “Timing of market entry – does it work?” from March 2019 simulates a popular crash timing strategy for the period 1970 to 2018. Our result is consistent with the consensus in the literature. One of the most important mistakes crash timers make is overlooking the statistically high opportunity cost of not investing before the crash.
Aspect 8: Cost-Averaging Effect
A topic that is related in terms of content to the question of SPE versus GEP and in some ways even identical is the well-known one Cost Average Effect (CAE), in German “average cost effect”. It is often mentioned in connection with fund savings plans. According to this effect, the mechanical, regular investment of a fixed amount of money, e.g. B. 200 euros per month in stocks have an advantageous return effect compared to his basic alternative of “investing everything immediately”. It would be nice. The CAE was identified over 40 years ago and since then in over a hundred scientific studies as a mirage, an error in thinking or method (first Constantinides 1979). Some time ago we discussed the reasons and manifestations of the CAE error in our blog post “The legend of the cost averaging effect” summarized.
Point of view 9: It never feels like the right time
For most private investors, the SPE versus GEP issue has almost nothing to do with how well or poorly the stock market performed in the previous four weeks, six months or five years. Is felt always the wrong time to start. If the stock market has risen significantly in the recent past, we fear an overvalued market that is in danger of crashing. If the market has fallen sharply in the recent past, we are concerned about the risk of the market collapsing even further. (Hence the hackneyed stock market quip “don’t grab a falling knife,” which the successful Berlin real estate investor and multimillionaire Rainer Zitelmann calls the “stupidest of all stock market sayings".) If the market has been sloshing back and forth for a long time without a very clear direction - such as in the last five months - then many believe that you have to wait until it has "calmed down" again. If you hear a private investor say “I would like to hold off on entering the stock market for now and wait – six months ago I wouldn’t have had these concerns and would have jumped in straight away,” then that is typically self-deception. In fact, our procrastinator always has or would have the entry problem. No matter how the market developed in the recent past. His real barrier to entry is not the current state of the market, as he imagines, but his very own psychology. Peter Lynch, one of the very, very few fund managers who have significantly beaten a correctly chosen benchmark over more than ten years, once said in this context: “Investors have lost far more money preparing for or anticipating market corrections than in the market corrections themselves.”
Viewpoint 10: Waiting for better times
An investor, let's call him Harry, is also very worried about whether the current entry point is particularly risky. However, Harry doesn't want the SPE or GEP method, but instead opts for BAW (wait until further notice with no time commitment to start). Harry feels that even an extended entry (GEP) over twelve or 36 months is still too risky. The BAW method is probably the one that around 95% of all Germans have always practiced. This method typically has one of two outcomes: either Harry never enters for the rest of his life, or he enters as a "performance chaser" (rearview mirror investor) a few years later near a stock market peak. Calculated from this point on, the returns in the years and decades after that will probably be below average. Those who never get involved will miss out on the long-term returns on the stock market that are seven times higher than those on savings accounts for the rest of their lives - like the vast majority of all households in Germany. We think the BAW method – if you even want to call it a “method” – is the worst possible choice of all. In order to be able to calculate your probably poor average return, you would first have to define this non-method in more detail, which we will save ourselves here.
Aspect 11: Irrelevance of the question in many practical investment situations
Finally: In many practical cases, the question of when to start does not even arise. The most important are the following constellations: (a) The money in question has already been invested in stocks and should now simply be invested differently (“new”). Here the investor is actually already in the market; In this respect, there is no entry that you have to or could time. (b) The money in question (or a relevant portion) is intended to flow into bond investments. There is no significant volatility in short- and medium-term high-quality bonds, which is why short-term timing here would be a waste of time either way (nota bene: this does not necessarily apply to speculative low-quality bonds or long-term bonds in general). (c) Anyone who has cash liquidity in a bank account and the amount exceeds the state deposit protection of 100,000 euros per customer-bank combination should not actually have to think about it as an informed, rational person: This Bank failure risk for the amount above 100,000 euros is so high that the pure timing risk of entering a globally diversified stock portfolio pales in comparison. (d) The stock investment amount in question is small relative to the investor's total assets, including his human capital, e.g. E.g. less than 10%. Here it probably doesn't matter what the return difference between SPE and GEP is in the end. He will be the speedometer needle at In totalThe investor's assets hardly move anyway.
Conclusion
Anyone who considers themselves to be a very rational homo economicus when it comes to finances will invest an existing amount of cash in the stock market at any given time as quickly and as quickly as possible, because in doing so they maximize the statistical expected return.
Anyone who considers themselves to be among the large majority of real existing Homo Sapiens who also have emotions when it comes to money will, in most stock market situations, be better off mentally and emotionally by entering the market with an extended amount of cash to invest and based on rules, e.g. B. over twelve or 24 months.
Doing nothing at all and waiting until the “right” or “better” time to start may be the worst and most common strategy at any given time.
Endnotes
[1] We cheated a little on this calculation. Over this long period, monthly returns for the global Stock market (i.e. 20+ countries) was not consistent for the first few decades. Therefore, for data availability reasons, we used monthly returns of the US stock market for the period 1926 to 1969 (CRSP 1-10 index) and monthly returns of the world stock market (MSCI World Standard Index) for the period 1970 to the end of May 2020.
[2] For reasons of simplification, at GEP we assumed a real interest rate of zero for the uninvested part of the portfolio; i.e. for the part that is only gradually invested over the course of 36 months and remains until then, for example, as a sight deposit in a bank account. In fact, over the last 120 years in the USA, this real interest rate has averaged 0.8% p.a., i.e. slightly above zero. However, our approach is unlikely to have any or only an insignificant influence on the overall result presented for two reasons. Firstly, we have not taken transaction costs into account anywhere, which would be much higher for GEP with 36 individual investments than for SPE. Second, today the real return on demand deposits is negative (i.e., worse than zero). This situation could continue for a long time.
literature
Constantinides, George (1979): “A Note on the Suboptimality of Dollar-Cost Averaging as an Investment Policy”; In: Journal of Financial and Quantitative Analysis; 14; June 1979; pp. 443–50.
Jacobs, Heiko; Weber, Martin (2016): “Random Walk Plus Drift – What Stock Prices Really Are”; Research for Practice, Volume 28, 2016; Behavioral Finance Group, University of Mannheim.
Kommer, Gerd (2017): “The Legend of the Cost Averaging Effect”; Internet location: https://www.gerd-kommer-invest.de/legende-vom-cost-averaging-effect/
Kommer, Gerd; Schweizer, Jonas (2018): “The underestimated risk of bank deposits”; Internet reference: https://www.gerd-kommer-invest.de/risk-von-bankguthaben/
Kommer, Gerd; Weis, Alexander (2019): “Timing the market entry – does it work?”; Internet reference: https://www.gerd-kommer-invest.de/timing-des-markteinstiegs/