The Dollar Cost Averaging strategy is quite flexible in the respect that many other strategies can use it as a basis for improvement. I created the InvestMete strategy through backtesting, and over the years it has shown to be very effective.
I believe that I created the InvestMete strategy in 1998. Looking though the history of purchases in my spreadsheet I can see that I normally sent $50 to my dividend reinvestment program of choice. Then in late 1998 the amounts changed to numbers like $71, $45, $66. This was because of InvestMete, but I need to step back to explain this.
Dollar Cost Averaging
Dollar cost averaging is a strategy where one breaks up their eventual stake in a company into numerous purchases over a period of time. In lieu of throwing a large amount of money into a position all at once, the idea is to make multiple purchases at different times. This means that portions of the position will be acquired at different entry points, thus reducing the risk that a single purchase might be made at the wrong time.
However, there is still a question as to exactly when the purchases will be made. If one has made the decision to break a large purchase up into ten parts then there is still the question as to when those ten purchases will be made. And that can lead to a different type of problem.
As humans we are subject to overconfidence and intuitive thinking. This will be the subject of an article in the future, but studies have found that we generally tend toward being opinionated about things that we are not well informed about – not exactly a big surprise. Additionally, hindsight bias can give us the impression that we are able to successfully make decisions intuitively.
This can lead to attempts to time the market – looking to purchase at the exact bottom of a dip and holding off on purchases when the stock is boasting a high. Too often we remember the few correct calls and forget the many misses. One can test their ability to make these decisions by creating a faux portfolio on Google and attempt to make such purchases and sales, but few do this over a long period of time.
Blind Dollar Cost Averaging
It is probably best to surrender to the fact that even seasoned professionals fail in attempts to properly time the market. Making blind purchases removes the temptation of timing – this is called Blind Dollar Cost Averaging.
In our example of making ten purchases, the only real decisions to be made are over what period of time this is to be done and how many purchases to make. If one decides to make ten purchases over the course of ten months then the first day of each month could be the day to make each buy, regardless of price.
As one who has made hundreds of purchases through dividend reinvestment programs, this timing situation is removed from my control and taken care of by the plan. One sends a set amount of money to the plan’s transfer agent, and the agent makes the purchases of all clients at the same time. There is no ability to select a specific date. If one is working through a discount broker then regular dates in the future need to be pre-selected.
This system works well, and it is amazing to see the fortuitous and unfavorable purchases smooth themselves out over time. It is also comforting to have the stress of attempting to get things perfect removed from one’s mind.
The deal with blind dollar cost averaging is that it takes part of the decision-making from the process away from us. Everything is on autopilot and to participate one only needs to follow the direction they originally outlined.
INVEST%
As one with a higher tolerance for risk, I wanted to find a similarly automatic means of skewing my monthly purchases toward companies in my portfolio that were not doing as well, and away from the ones that were sporting new highs. I was fortunate to find an idea from Moneypaper to be exactly what I had been looking for.
Their system is called INVEST%™. The idea is that when a company’s current price is doing well in comparison to its past 52 weeks, less money than originally planned should be sent. When the current price has slumped more should be sent. If the company is currently at its 52-week high then 50% of the normal amount is sent, and if it is experiencing its 52-week low then 150% of the normal amount is sent. If the current price is somewhere in between these two extremes then the percentage would be set relative to the 52 week high and low range. Directinvesting has a service that can do the work for you.
A formula for figuring out the amount to send is fairly simple:
(1 - ((Current Price - 52WeekLow) / (52WeekHigh - 52WeekLow))) + 0.5, expressed as a percentage. Looking at an example illustrates this.
Aflac
- 52 week high = $57.18
- 52 week low = $46.96
- Current price = $51.82
Plugging these numbers into the formula we get:
(1 - ((Current Price - 52WeekLow) / (52WeekHigh - 52WeekLow))) + 0.5(1 – (($51.82 - $46.96) / ($57.18 - $46.96))) + 0.5Simplifying we get:
(1 – (4.86 / 10.22)) + 0.5(1 – 0.48) + 0.50.52 + 0.5 = 1.02 - expressed as a percentage = 102% In this example we can see that Aflac is slightly closer to their 52 week low than their 52 week high, so if the plan had been to make a $100 purchase of Aflac then that amount would be increased to $102. Had the current price equaled the 52 week low the plan would be to send $150, and if it had matched the 52 week high then $50 would be sent.
This makes plenty of sense to me but there are a couple of issues I had with it.
The first is that this is optimized for dividend reinvestment programs, as opposed to purchases made through discount brokers. There is nothing that can be done about that. In a DRiP one can send the $102 knowing that the entire amount will be used to purchase stock, including partial shares. Although discount brokers might use dividends to purchase partial shares, only whole shares can be purchased by the investor through their system.
In our example it is coincidental that the $102 would just barely purchase two shares, but what if the intention had been to make a $75 purchase? That would mean that $51.82 would go toward the purchase of a single share and the remaining $23.18 would go unused. As there is nothing that can be done about it (as is also the case with InvestMete) the best thing to do would be to save the $23.18 and add it to next month’s $75, so that one would have $98.18 for that next purchase.
The other issue is that if one is planning to make several purchases of companies that are not doing well then they will all require larger purchases than anticipated. If one has the money to invest then that is fine, but this is not always the case. I wanted a way to, for instance, plan to make a $250 purchase divided amongst four different companies and not have that total amount altered by the INVEST%™ formula. This is why I created InvestMete.
InvestMete
InvestMete calculates the INVEST%™ for each company, then metes the amounts to be sent relative to one another. Below is an example of some stocks in my portfolio.
Amount to be sent: $250.00
| Symbol | 52 Week High | 52 Week Low | Current Price | InvestMete | Amount | Shares |
|---|---|---|---|---|---|---|
| AFL | $57.18 | $46.96 | $51.82 | 31.26% | $78.16 | 1.51 |
| JNJ | $150.17 | $126.10 | $147.81 | 18.25% | $45.63 | 0.31 |
| MMM | $219.75 | $150.58 | $177.00 | 34.12% | $85.30 | 0.48 |
| WTR | $52.58 | $33.82 | $51.90 | 16.36% | $40.91 | 0.79 |
In this instance, the planned $250 has been properly divided into four amounts, adjusted relative to their INVEST%™ calculations. For instance, AFL and MMM are not doing as well as JNJ and WTR when compared to their 52-week high/low range, so a greater percentage of the whole is allocated to them. These four amounts would be sent to the DRiP plan administrator for the purchase of stock.
If one is using a discount broker then a problem is plain to see. Aflac would be the only stock that could be purchased. There is not enough planned money to cover the purchase of any of the other three stocks because that requires the ability to be able to buy partial shares.
One option would be, as suggested above, to purchase AFL and save the remaining money for the following month. This would mean $198.18 + $250 = $448.18 would be available.
A better option might be to simply not send to companies that are doing well. In this example the InvestMete is lowest with JNJ and WTR, as these two companies are closer to their 52-week highs than the other two companies, so removing them shows the following:
Amount to be sent: $250.00
| Symbol | 52 Week High | 52 Week Low | Current Price | InvestMete | Amount | Shares |
|---|---|---|---|---|---|---|
| AFL | $57.18 | $46.96 | $51.87 | 47.94% | $119.84 | 2.31 |
| MMM | $219.75 | $150.58 | $177.74 | 52.06% | $130.16 | 0.73 |
Sometimes this works, other times it does not, as the price of MMM still does not allow the purchase of a whole share. Perhaps in this instance, one could purchase a single share of each, which would allow $20.39 to be rolled over to the following month. It’s just a matter of deciding upon a workaround for this situation.
Of course, calculating these things can be very time-consuming so I have put together a couple of programs that make this task a bit more palatable.
If you have a Google account then you have got access to Google Drive and Google Sheets. I have created the page InvestMete for Google Sheets, which offers a link to the spreadsheet I use, as well as an explanation of how to put it into Google Drive for use. Unfortunately, this is not an Excel spreadsheet and only works through Google Drive, but perhaps someone will adapt it in the future and allow me to offer it here.
If you would rather simply use an online version then I have created the page InvestMete, where you can enter a series of tickers, each separated by a space, and enter the amount to be dispersed. The program will then calculate this information. For readers in Canada who know that Canadian stocks may have different ticker symbols on different exchanges, I would note that I use information from Yahoo Finance, so if a ticker doesn’t work then that would be the place to find the symbol that will work.
To Sum It Up
I used InvestMete for many years to charge my dividend reinvestment portfolio and am now using those dividends in my retirement. Not only did the strategy backtest well when I created it, but it worked well in the real world. The strategy works best for those who have their monies in dividend reinvestment programs so that they can send the exact amounts suggested. However, those using discount brokers can adapt the idea of the strategy in their own way to determine how much to invest.
This should be a consideration for any dividend investor who feels comfortable with adding a bit of risk to their conservative portfolio.
INVEST%™ is a Registered Trademark of The Moneypaper Publications LLC
