In Intelligent Investor, Ben Graham discussed two alternatives to market timing: rebalancing and dollar cost averaging. Rebalancing is a tool used to position a portfolio between asset classes; in Graham’s discussion, this was a range of 25-75% weighting in either stocks or bonds. Dollar cost averaging involves investing equal amounts of capital into a particular security at set increments (say, $1000 a month). The two strategies contrast one another in a fundamental way: rebalancing is an active strategy, while dollar cost averaging is passive. As someone closely follows the teachings of behavioral finance, this presents an interesting dilemma. Rebalancing is an effective strategy, except that it involves subjective decision-making, which brings emotions and behavioral biases into play while attempting to answer an ambiguous question. During times like March 2009, fear is running high, causing smart investors to miss fantastic opportunities; on the other end of the spectrum, when optimism hits, some companies trade at absurd multiples (100x revenues, like Facebook, Twitter, etc), suggesting that the internet bubble of a decade ago may be lost in the minds of many. As intelligent investors, it is important that we find a way to protect ourselves from these mental setbacks and stick to an effective investment strategy.
Dollar cost averaging, on the other hand, is passive and only requires the decision of overall security selection. For most people, this isn’t too difficult: if you told them they had to own the stock for the next 20-30 years, you would likely find a general consensus in large cap blue chips like JNJ, KO, PEP, and PG. With this strategy, the investor will buy the security when it is overvalued, undervalued, and fairy valued, and will end up with returns (over time) in line with the overall business. While that strategy would provide attractive returns, it undermines the active investors training and understanding of valuation. With this strategy, I’m locked into buying WMT at 50-60x earnings (like in December 1999), even when I know that they can never grow fast enough (from a $200 billion base) to justify that multiple. As an active investor with a long term holding period, my question is this: how can I utilize these strategies in a way that gets the best of both worlds? How can I avoid the inherently biased decision making of rebalancing, while using the strategy of dollar cost averaging in a value portfolio?
My response (a work in progress) to this question is an active approach to dollar cost averaging. The first point to be made is that these are Buffett-style investments: I want companies that I can hold forever, than have sustainable competitive advantages, and that will be bigger and better ten, twenty, and fifty years from now (the four mentioned above are good examples). From there, I want to differentiate between where I’m a buyer and where I’m inactive (or selling if that interests you); I’m advocating purchases equal to 5% of cash on hand when the price falls below 15x trailing twelve months earnings, 10% below 12x, and accumulating capital at any multiple above that. Assuming someone saves $1000/month for investing, let’s run through an example for Johnson & Johnson. (Assumptions: I’m going to use yearend earnings and start on 1/1/2007; any P/E’s for a given year’s purchases are TTM; all purchases are based on trading price from first day in new month).
In 2006, diluted EPS was equal to $3.73/share; at fifteen times earnings, we are willing to buy at $55.95, which the price never dipped below. In 2007, diluted EPS was $3.63; while it got close, the price never fell below 15x earnings ($54.45) in 2008 either, so we are holding $24000 in cash at the end of the year.
In FY 2008, EPS was $4.57/share, indicating that we are buyers below $68.55/share; starting January 1, 2009, we are buyers every month of the year. On January 1st and February 1st, we buy positions equal to 5% of our cash position, or $1250 in January, and $1237.50 in February. In March ($2451.25), April ($2306), and May ($2175.5), we are buying 10% of cash on hand (since it dipped below 12x TTM), followed by 5% for the remainder of the year. By implementing this strategy, the cost per share for the year is equal to $55.78 per share ($16,595.24/297.52). The purchases in 2009 are broken down here:
To avoid bogging down in the numbers and calculations, I will simply reprint the annual results:
For the entire period, you would have spent $33,363 on the purchase of 541.93 shares. The cost basis per share from 2007 to date using this strategy would be $61.56/share. Even without number crunching, we know the results are better than simple dollar cost averaging because we purchased more (10% of cash) in March, April, and May 2009, when the price was below $50/share. We also avoided buying in the mid-60’s and low-70’s in 2007 and 2008, when the multiple exceeded 15x earnings. Essentially, the strategy is active management, but based on passive, objective valuation metrics. This same idea can be applied on the portfolio as a whole: you could find market metrics (like the 10 year trailing P/E on the major indices, and Treasury bond yields) to compare stock and bond valuations to set your weighting between asset classes, and adjust them as the need arises.
One can easily stop here and think, yeah, that makes sense, but what about all the recalls? And in March 2009, when you were buying 10% of your cash on hand, what about the macroeconomic picture? That is exactly the point of this strategy: when the world is falling apart and uncertainty is rampant, our biases takeover and we freeze. As Warren Buffett of Berkshire Hathaway (BRK.B) has noted before, you pay a pretty penny on Wall Street for a cheery consensus. With this strategy, we can remove ourselves from the noise and invest in companies with strong long term fundamentals at bargain bin prices.
This is a work in progress, with simple rules based on a less than perfect metric (P/E). As I sit down and develop this further, adjustments will be made to strengthen the requirements, and the overall strategy for purchases and holding periods will become more in depth. For example, share purchases below 12x earnings (and at 10% of cash) were only made in three months out of the entire test period; this suggests that this number could be adjusted as a higher percentage of cash (say, 15-20%) to take advantage of these types of opportunities. An investor could also break the strategy down further, and have set rules at 8x, 10x, etc, which would create larger purchases during March 09’ type opportunities and drive down average cost per share. Regardless of the methodology used, the point are clear: by setting up a strategy that is based around dollar cost averaging and objective valuation metrics, an investor can take advantage of the benefits presented from both active and passive investment strategies within their individual portfolio.
Dollar cost averaging, on the other hand, is passive and only requires the decision of overall security selection. For most people, this isn’t too difficult: if you told them they had to own the stock for the next 20-30 years, you would likely find a general consensus in large cap blue chips like JNJ, KO, PEP, and PG. With this strategy, the investor will buy the security when it is overvalued, undervalued, and fairy valued, and will end up with returns (over time) in line with the overall business. While that strategy would provide attractive returns, it undermines the active investors training and understanding of valuation. With this strategy, I’m locked into buying WMT at 50-60x earnings (like in December 1999), even when I know that they can never grow fast enough (from a $200 billion base) to justify that multiple. As an active investor with a long term holding period, my question is this: how can I utilize these strategies in a way that gets the best of both worlds? How can I avoid the inherently biased decision making of rebalancing, while using the strategy of dollar cost averaging in a value portfolio?
My response (a work in progress) to this question is an active approach to dollar cost averaging. The first point to be made is that these are Buffett-style investments: I want companies that I can hold forever, than have sustainable competitive advantages, and that will be bigger and better ten, twenty, and fifty years from now (the four mentioned above are good examples). From there, I want to differentiate between where I’m a buyer and where I’m inactive (or selling if that interests you); I’m advocating purchases equal to 5% of cash on hand when the price falls below 15x trailing twelve months earnings, 10% below 12x, and accumulating capital at any multiple above that. Assuming someone saves $1000/month for investing, let’s run through an example for Johnson & Johnson. (Assumptions: I’m going to use yearend earnings and start on 1/1/2007; any P/E’s for a given year’s purchases are TTM; all purchases are based on trading price from first day in new month).
In 2006, diluted EPS was equal to $3.73/share; at fifteen times earnings, we are willing to buy at $55.95, which the price never dipped below. In 2007, diluted EPS was $3.63; while it got close, the price never fell below 15x earnings ($54.45) in 2008 either, so we are holding $24000 in cash at the end of the year.
In FY 2008, EPS was $4.57/share, indicating that we are buyers below $68.55/share; starting January 1, 2009, we are buyers every month of the year. On January 1st and February 1st, we buy positions equal to 5% of our cash position, or $1250 in January, and $1237.50 in February. In March ($2451.25), April ($2306), and May ($2175.5), we are buying 10% of cash on hand (since it dipped below 12x TTM), followed by 5% for the remainder of the year. By implementing this strategy, the cost per share for the year is equal to $55.78 per share ($16,595.24/297.52). The purchases in 2009 are broken down here:
| Starting | Cash Balance: $24,000 | Shares Purchased: 0 Shares |
| Jan and Feb (5%) | $21,512.50 | 39.8 Shares |
| March, April, May (10%) | $19,579.60 | 136.66 Shares |
| Last 7 Months (5%) | $19,404.76 | 121.06 Shares |
| Year End 2009 Holdings | CASH: $19,404.76 | OWNERSHIP: 297.52 |
To avoid bogging down in the numbers and calculations, I will simply reprint the annual results:
| 2010 Purchases (EPS: $4.40) | Year End Cash: $19,218.31 | Shares Purchased: 197.1 |
| 2011 Purchases (EPS: $4.78) | April 2, Cash: $18,636.31 | Shares Purchased: 47.31 |
| ENDING POSITIONS | TOTAL CASH SPENT: $33,363 | OVERALL OWERSHIP: 541.93 |
For the entire period, you would have spent $33,363 on the purchase of 541.93 shares. The cost basis per share from 2007 to date using this strategy would be $61.56/share. Even without number crunching, we know the results are better than simple dollar cost averaging because we purchased more (10% of cash) in March, April, and May 2009, when the price was below $50/share. We also avoided buying in the mid-60’s and low-70’s in 2007 and 2008, when the multiple exceeded 15x earnings. Essentially, the strategy is active management, but based on passive, objective valuation metrics. This same idea can be applied on the portfolio as a whole: you could find market metrics (like the 10 year trailing P/E on the major indices, and Treasury bond yields) to compare stock and bond valuations to set your weighting between asset classes, and adjust them as the need arises.
One can easily stop here and think, yeah, that makes sense, but what about all the recalls? And in March 2009, when you were buying 10% of your cash on hand, what about the macroeconomic picture? That is exactly the point of this strategy: when the world is falling apart and uncertainty is rampant, our biases takeover and we freeze. As Warren Buffett of Berkshire Hathaway (BRK.B) has noted before, you pay a pretty penny on Wall Street for a cheery consensus. With this strategy, we can remove ourselves from the noise and invest in companies with strong long term fundamentals at bargain bin prices.
This is a work in progress, with simple rules based on a less than perfect metric (P/E). As I sit down and develop this further, adjustments will be made to strengthen the requirements, and the overall strategy for purchases and holding periods will become more in depth. For example, share purchases below 12x earnings (and at 10% of cash) were only made in three months out of the entire test period; this suggests that this number could be adjusted as a higher percentage of cash (say, 15-20%) to take advantage of these types of opportunities. An investor could also break the strategy down further, and have set rules at 8x, 10x, etc, which would create larger purchases during March 09’ type opportunities and drive down average cost per share. Regardless of the methodology used, the point are clear: by setting up a strategy that is based around dollar cost averaging and objective valuation metrics, an investor can take advantage of the benefits presented from both active and passive investment strategies within their individual portfolio.
