Long Term Investing
My prior paper, position sizing, discussed a few techniques for getting into a stock for short term trades. It concluded with defining an “R,” your risk per trade, and your capital risk. From R and capital risk, you determine a position size and enter the stock in one lump sum.
If the stock hits your stop, you exit. Otherwise you manage the trade and determine an exit later.
This is perfect for short term trading. What about long term investing? Investing when you are contributing a set amount of money each month into your retirement account. In this situation you really don’t have a stop loss. You are now a long term investor.
While there are many ways and systems for getting into a stock, I want to discuss three: Constant share (CS), dollar cost averaging (DCA), and value averaging (VA).
These next techniques all divide your capital into parts and attempt to capture lower prices and thus bring your average cost down. No stock can go up every day and no stock can go down every day. With any type of averaging down in a stock going down you have to ask yourself which position you take: “If you liked the stock at $50, you should love it at $40” or “trying to catch a falling knife.”
Essentially, these are “buy low” and “buy higher” strategies. Most of these do not have any type of exit or sell plans.
Constant Share (CS) – fixed number of shares
With this strategy you buy a set amount of shares each time period. So, every month or quarter you buy 10, 20, or 100 shares of XYZ corporation. The amount you invest will change with the price of the stock. One month it may be $20 and your 10-share investment is $200. Another month it is $18 and you contribute $180 to your retirement and the next it is $22 and you sock away another $220.
Dollar Cost Averaging (DCA) – fixed amount of money
Here you invest a fixed amount of money, say $200, at regular intervals – weekly, monthly or quarterly.
DCA is like an annuity. You put a fixed amount of money in every month. When the stock goes down you are able to buy more shares. When the price goes up, you buy fewer shares.
Value Averaging (VA) – fixed growth rate, variable number of shares and money
Like DCA, this technique aims to invest more when the price falls and less when the price rises. Instead of a “fixed dollar” amount each month VA wants the portfolio to go up by a certain amount each time period. One major difference with DCA is that when the market has a nice upswing you may actually sell shares instead of purchasing more. Remember, you want the portfolio to go up a certain amount. If that amount is exceeded by a rising stock value, then you get to sell shares and pull out some money.
You set a target growth rate; the amount you want your portfolio to go up each month. If the stock has gone up, then the stock is contributing part of this growth rate and you only need to invest the difference. When the stock goes down then you have to contribute more making up the shortfall to maintain the growth rate.
CS vs DCA vs VA
Long term investing means years or decades. In the longer term, VA suffers the same problem as DCA in that it fails to take market growth and inflation into account in its “linear value path” (adding the same amount to value every month). CS has you purchasing the same number of shares year after year. During this time, inflation and the value of the stock is increasing (approximately 10% annually) and you are investing more during each investment period.
Not true with DCA and VA. You are adding the same amount in a constant linear manner while the value of the stock continues to rise. After a few years you are probably taking money out several time periods with VA. This means you have less money working for you in the market and more in cash doing nothing.
An example from Edleson’s book, “Value Averaging: The Safe and Easy Strategy for Higher Investment Returns,” compares CS, DCA, and VA if you followed each system to the tee from 1926 to 1991.
CS keeps up with the market over time. The incremental investment increases as the share price increases. DCA buys fewer and fewer shares over time as the price of the stock rises. With VA, shares are subject to being sold. Because of this neither DCA or VA can reach a total portfolio value (#shares x price/share) like CS. In fact, with VA you are actually selling and trimming shares along the way. You are pulling your money out as you go instead of investing more.
In the beginning you invest a lot, say $100/month, but eventually the increase in value gets you to the point where you are not adding anymore and occasionally taking money out. CS is constantly purchasing more.
The 1926 to 1991 portfolio readily exposes this problem. The VA final value (remember you are taking money out along the way) is $8,220 versus more than $422,000 for CS.
The two graphs below show a comparison between DCA and VA.
The system can be tweaked with “growth equalization” meaning that the total invested keeps up with the market and CS. If the market grows at a certain rate, you are spending more with CS. If the annualized growth of the market was 12% (1% month) then you increase your monthly investment by this amount (for both DCA and VA). Doing this results in a portfolio value, roughly equal for all three, of $420,000.
Finally, what frequency of investing should be done? Daily? Quarterly? Yearly?
This is timing. It is extremely difficult to time the market. If it goes up one day we don’t want to sell because it probably will go up again the next day. When it is going down, the farther down it goes the better our price. So, what timing is best?
This chart shows that Quarterly is best. But remember, all these numbers are for a long-term investment plan of several years in the same stock.
A final nail in the coffin for DCA.
In 1994, Marshall and Baldwin published, “A Statistical Comparison of Dollar-Cost Averaging and Purely Random Investing Techniques.” Random is defined as the same amount of money as DCA, $400/period. But there is a 50% probability of investing in a particular period and an equal chance of investing either 150% or 250% of the amount invested with DCA. This mimics a typical investor “on/off” or “more/less” investment pattern. Also, the expected value is the same as in DCA. What they found was that with a 99% confidence there was no significant difference between DCA and random investing.
Marshall repeated this study adding VA, “A Statistical Comparison Of Value Averaging Vs. Dollar Cost Averaging And Random Investment Techniques” in 2000. The basic premise being that, according to the efficient market hypothesis (EMH) few, if any, techniques should be superior. Marshal ran twelve 500 run simulations. Each time period was 5 years with a purchase each year.
In the following tables we compare VA to DCA. The amount invested for VA is increased 10% annual to reflect market growth.
As you can see, VA beat DCA in all market conditions, uptrend, downtrend and sideways. It had both lower average share costs and a higher internal rate of return.
Marshall also explored the effect of market variability quarter to quarter. Does it matter if the variability of the market changes by 1% or 25% each quarter? What if your investment horizon is 2 ½ years or 10 years?
And it did not matter the market direction, up market (favorable), down market (unfavorable), or sideways market (neutral).
Long Term vs Short Term Investing
As you can see, VA is by far the most efficient method of long-term periodic investing. This is the type of investing you do for years to build wealth when you are contributing on a regular basis to your retirement or savings.
What if you have a large lump-sum of money that you want to invest? Do you go all in, lump sum, or do you add periodically like VA, leaving a sizable amount of money in cash waiting for its turn to be invested and go to work?
Seth J. Masters of AllianceBernstein did a historical analysis of the US stock market from 1926 to 2013. This encompasses over 1,000 different entry points across a range of market environments, from the Great Depression to the bull markets of the 1980s and the global financial crisis in 2008.
Lump sum investing was better. Statistically the market grew 12.2% during this time while cash only grew 3.6%. When you DCA you are holding a significant portion of your capital in cash missing out on the market gains. In this study Seth looked at one year periods of investing in a lump sum at the beginning vs DCA contributing monthly for 12 months. LS investing was clearly superior. He did not compare LS to VA.
Why is lump sum investing so important. Because you don’t miss days of rapid market increases. Fidelity investments shows us in the graph below just how much missing those few days can cost us.
Peak to Peak
Jason Hull, CFP, asked the question what if you bought right at the peak of the market on October 11, 2007 (1,554.41) and using various investing techniques held your stock until the market fully recovered on March 7, 2014 (1,878.04). If you were really, really good at market timing you could have taken the money out and put it in a CD earning 0.5%. You wouldn’t have lost money, but you would never really make money.
- Lump Sum: if you invest $100/month for the entire 77 months you would have $7,700 invested. So, lets invest all of it on Oct 11,2007 and enjoy the ride.
- DCA: investing $100/month
- VA: increasing the target value by $100. (as learned above, VA is best used with an inflation or growth factor such as 7% or 10%. This study used straight VA.
- VA: with an added growth factor. Jason used the S&P 500 CAGR (Compound Annual Growth Rate).
But remember, Jason explored the worst market timing in recent memory. Lump sum had to sit and wait over 70 months to get back to even. That’s why an averaging system beat it so handily.
I would be remiss if I didn’t explore one last method, one favored by gamblers the world over, Martingale.
This was popular in 18th century France, it is a double down system. Take a coin-toss, for example, with a 50:50 chance of heads. You bet heads for a $1 and lose. Just double up and bet $2, then $4, then $8 until you eventually win. If you have infinite wealth, and your opponent does not set a betting limit, you will eventually win.
A variation of this is practiced by some stock newsletters. There is never a preset amount to invest and you hopefully have unlimited capital. Say you buy 100 shares of a stock at $100 for $10,000. The stock drops 10% to 90 so you double down. Now you own 200 shares at an average price of $95 for a total of $19,000 leaving you 5% down instead of 10% down.
The stock drops again, this time to $85.5, and your position is once more down 10% from your cost basis of $95. You double down, buying 200 shares at $85.50, yielding 400 shares at an average price of $90.25 and an investment of $36,100.
Once more it drops to $81.225, you double down again buying 400 shares for a total of 800 shares with an average price of $85.74. Your total investment is now $68,590. Just like doubling down at roulette this will eventually ruin you unless you have Bill Gates money. The other disadvantage is that you must have this capital readily available sitting on the sidelines. If you are good, and every investment makes 10% before it loses 10%, you wind up having a sizable amount of money in cash waiting – doing nothing!
It is hard to beat lump sum investing. You don’t want money sitting in cash doing nothing and missing out on the best days of the market while the market increases year after year. But, if you don’t have a million in cash laying around to invest, you need to build your investment gradually. In this case I have shown that the best method of incremental increases is value averaging adjusted with a growth factor.