The member-only community forum (inactive today, as we’re moving to a Premium-member Facebook group) is an invaluable source for learning and receiving feedback and support. Members bring forth questions and topics not commonly addressed by traditional blogs and services, or those for which the web is filled with too much contradicting information.
We are a small and exclusive community, and thus each posting member gets the maximum personal attention to resolve any investing question or dilemma*. Surfacing questions and doubts on the forum, not only helps members get quick support and resolution, but also helps me evolve the service in those aspects that are of interest for our members.
Recently, a member asked the following question (here’s a link, make sure you are logged in to view it):
“Do you have any recommendations on how the QV strategy could be implemented with accounts receiving monthly deposits such as the IRAs I manage? How could the “new money” be integrated, either at the yearly rebalance or any other time? I have considered adding new positions or adding to existing positions, or ???? In this case taxes are not an issue. Thanks!”
The question is an excellent opportunity to bring to discussion the topic of Dollar-Cost Averaging.
I am happy to delve into this concept and provide an elaborate answer on this topic, which may be of interest to many of our members.
Dollar-cost averaging (DCA) is the practice of investing a fixed dollar amount in a stock or a portfolio of stocks every single month. If, in a certain month, stock prices decline, then the fixed dollar amount buys more units of that stock/portfolio. If in another month, stock prices appreciate, then the same fixed dollar amount buys fewer stock units. Such a mechanical formula guarantees that you buy more stock units when prices are lower, and buy fewer units when prices are higher. In its essence, the mechanism averages the purchase price over time. Here is what Benjamin Graham has written about dollar-cost averaging in his book, The Intelligent investor (p.118):
Earlier in the book, Graham discusses a case study, where the market declined during a 20-year period but an investor using dollar-cost averaging managed to make 8% annually (The intelligent Investor, p. 2):
But that is not always the case. Dollar-cost averaging is not always better than investing a lump sum in the beginning. In some cases, it results in lower returns.
The general rule is, when the stock market declines during the initial period of implementing a dollar-cost averaging scheme, its performance is better than the alternative of lump-sum investing at the beginning of the period. By implementing dollar-cost averaging, the buying price is averaged down during the all-important initial buys. However, when the stock market is on an upward trend, delaying purchases by using dollar-cost averaging, will only result in adding shares at higher prices. The results in such case would naturally be inferior to investing the entire amount in a lump sum at the beginning. When the market goes sideways, with multiple up and down runs, it is harder to determine which of the two is superior, lump-sum investing or dollar-cost averaging.
While Graham’s advocacy of DCA was considered undisputable for many years, during the past few decades, several critics emerged. Here is a good summary of a Vanguard paper, concluding that both the stock market’s history, as well as investment theory, supports immediate investment (a.k.a lump-sum investing). Additional resources supporting the case for lump-sum investing are: 1) An excellent article by superstar financial advisor Michael Kitches 2) A 2001 academic paper by Leggio and Lien.
Dollar-cost averaging is considered by some to merely be a cognitive fallacy. Other than being a method for achieving better risk-adjusted returns, it is regarded as a method to minimizing regret. Since human beings give a higher weight to losses than they do for gains, dollar-cost averaging reduces the probability of entering the market in the “wrong” price. People may feel intensified feelings of regret if they realize that they’ve entered the market at the wrong price (when a decline follows), than if they average their entry points, even if they don’t earn the maximum potential gain. That may be irrational, but it is how our minds are wired.
Could it be that Benjamin Graham, a master of portfolio construction and behavioral investing, recommend a sub-optimal investment scheme? Wasn’t he aware that dollar-cost averaging is based on a cognitive fallacy?
Well, I believe that I figured out the mystery. I believe that the popular discussion on dollar-cost averaging is tainted with some inaccuracies. Allow me to attempt to clarify.
First, we must distinguish between discretionary dollar-cost averaging (one possess the funds to invest a lump sum but chooses to delay some purchases) and involuntary dollar-cost averaging (new funds for investment become available every month).
Let’s start with the former case, where an investor has a sizable lump-sum available for investment, and she has the choice whether to deploy the entire amount in bulk or keep most of it in cash, deploying only a portion of it every month. In such a case, the investor has to ask herself: What kind of an investor am I? Am I after the maximum return potential, while I know that the path to get there may be volatile? Or am I willing to forego some of the potential returns if I could limit the drawdown and volatility of my overall portfolio? If the answer is the former, then I would suggest sticking with lump-sum investing, and not attempting to dollar-cost average. The Vanguard study should be considered – the probability of dollar-cost averaging to underperform lump-sum investing is higher than 50%. Nevertheless, for conservative investors that are not chasing the absolute highest returns, but rather, seek adequate return coupled with peace of mind – my suggestion would be to engage in dollar-cost averaging. history shows us that dollar cost averaged portfolios are less volatile and incur lower drawdowns.
Recall Graham’s definition of investing vs. speculation:
“An investment operation is one which, upon thorough analysis, promises safety of principal and a satisfactory return. Operations not meeting these requirements are speculative.”
Graham intentionally used the term “satisfactory return.” He didn’t use the term “highest returns” or “best returns,” but merely satisfactory returns. having read Graham’s work thoroughly, I honestly believe that Graham was deliberately willing to yield some of the returns in exchange for having a less volatile portfolio which provides investors with peace of mind. It seems like a sensible approach helping humans hold on to their portfolios for longer.
Thus, I believe that Graham was not a victim of a fallacy. I believe he was fully aware of the empirical evidence showing that lump-sum investing has a higher likelihood to perform better. Yet his assumptions on the utility function of investors were different from the critics who followed him.
Let’s know tackle involuntary dollar cost averaging, which is a very common case. Many investors wish to allocate a portion of their monthly salary or income to an investment Portfolio. How would an investor then allocate his capital? This would be a case where there is no choice between lump-sum investing and dollar cost averaging. The question becomes – What would be the best practice to build a quantitative portfolio gradually over time with new funds which become available every month?
That was the exact question that Carol raised in our members’ forum.
And that is what I responded to her.
“Building your initial portfolio:
First, decide on the number of positions you want to hold and the minimum position size you allow. The latter should be considered per the minimum transaction costs you pay. I recommend not exceeding 0.2%.
Let’s say that you have $2000 of new money every month. Let’s also assume that you wish to build a 50-stock portfolio (30 QV, 20 QM, each position 2% allocation). You invest through interactive brokers who charge $1 minimum per transaction. In that case, your minimum position size will be $500 (1$ transaction costs / 0.02% = $500).
Thus, every month you would buy 4 new stocks with the $2000 of new money. It will take you 13 months to build a 50-stock portfolio.
Use the screeners to choose the stocks to invest in. Choose the stocks that rank the highest (follow the instructions carefully), and that you don’t already own.
Maintaining and adding new money to your portfolio:
Perform rebalancing according to the recommendations of every screener – QV every 12 months, QM – every 6 months. Rebalance before you add the new money that came in that month.
Don’t rebalance every month and don’t rebalance stocks that you held less than the recommended holding period.
Invest the new money that comes in the following way – Add to the positions which are most compelling at that time. In our example, keep the minimum transaction size to $500, to minimize transaction costs. Add $500 for the four most compelling positions you already hold.
How do you determine the most compelling positions? See who ranks the highest in the screeners (following the screener instructions carefully).
The benefits of the proposed method above:
The number of positions are contained and do not grow indefinitely. Portfolio maintenance is easier that way.
Rebalancing is done at the prescribed times, which reduce transaction costs.
You add new money to the most compelling positions at that time.”
Holding value stocks for a longer period than a year is perfectly fine. Therefore, rebalance a value portfolio a year after completing its buildout, as long as the buildout time is reasonable (let’s say, 3-12 months).
As an example, let’s assume that it takes you 6 months to build your value portfolio to the desired number of positions. Do the first rebalance 12 months after the last set of stocks were added, which is 18 months after the first set has been purchased. Then, keep rebalancing every 1 year.
Momentum stocks are different since their momentum wanes as time progresses. I recommend building a momentum portfolio at once, in a lump-sum, and not scaling in. You can add new money every rebalances period, which is typically 6 months.
As for new money added to existing positions, they should not affect the rebalancing period. If you need to sell a holding during a rebalance, sell all of it, even if you added to it only a month ago.
The rationale for these rules – keep it simple and act rationally. I’m sure that wise investors will come with various optimizations, which I would willingly adopt if they simple enough and rational.
*Due to regulatory restrictions, we avoid providing personalized investment advice, and prohibit any questions which are personal in nature.