Dollar Cost Averaging - The Basics, the Advanced, and Where it breaks down

Matt

Administrator
Staff member
Dollar Cost Averaging is a quite well known method of investing that avoids the problems of trying to time the market. The basic premise is to add funds into an investment over a period of time.

Frequently the term is used to describe a long term savings plan, such as contributions to a company 401(k) plan that are taken from your paycheck on a monthly basis. An example 401(k) might have funds split into a basic 2 class structure of stocks and bonds, 80/20 between them. If you were to have a $1400 monthly payment going into the funds, each month they would buy:
  • $1120 of Stock funds
  • $280 of Bond funds
Many such plans would allow for fractional ownership. Today, with a price of $101, the $1120 would buy 11.089 shares of VTI. If next month the fund appreciated by 5% to a new price of $106.50, the same $1120 would buy 10.516 shares.

The concept here is that with a consistent dripping of the investment your basis will 'average out' and you will be able to enter the market without the fear of buying high. The 401k style of Dollar Cost Averaging is more accurately known as a periodic investment strategy.

Dollar Cost Averaging Vs Lump Sum investing.

Dollar Cost Averaging is also a strategy used to drip a lump sum (rather than income) into the market over time. By setting a fixed period of time to invest it is thought that you would get in at a better rate. This is highlighted by this graphic by Vanguard:

Dollar Cost by Vanguard.PNG

This image, and variants of it are lauded by many financial firms as how DCA is a savvier financial decision, however it is actually a very biased data set, and used solely here to show how DCA works, when it works (IE should the price of later serial payments have decreased) if instead the value of the investment has increased then DCA is inferior to lump sum contributions.

So, in some ways you still have a game of guess work, but it does help facilitate a decision to invest when there is the fear of market fluctuations at play.

Should fear drive your investments?

Interestingly, the notion of DCA plays heavily upon marketing timing and fear. While it seems to present itself as a mitigation tool regarding such things, one could argue that simply acknowledging market timing is to fall under its spell. There is a very interesting article, also presented by Vanguard that proposes that DCA fails against Lump Sum Investing over three markets (US, UK and Australia) across a wide time frame: PDF Here

Chart from Vanguard

DCA vs LSI.PNG

You have to get in at some point, the notion of drip feeding a lump sum comes with the downside of not having the balance invested in a meaningful investment. Furthermore, with each investment, as money moves from outside of the asset to inside of the asset the ability for the DCA to offset fluctuation decreases. And once the final DCA investment is made you have then lost all power, you are effectively in fully, and you no longer have the benefits of DCA going forward.

Repair Ratio
There is a ratio between a singular investment and the ability to repair it to par in the event of decline. The notion of DCA works along these lines, but eventually declines, and at this point investments need to shift from the core asset to a defensive asset class. Correlation is very important at this point, there is no value in selecting two assets that react in tandem to similar events if you are seeking to embed repair value to a portfolio. This is the premise behind diversification of assets, and it is something that you can explore in the comments, and I will extrapolate on in future posts. For now, let it suffice that diversification of assets too soon might be detrimental, but at a certain ratio of assets to income a repair variable should be implemented.
 

SC Parent

Level 2 Member
This image, and variants of it are lauded by many financial firms as how DCA is a savvier financial decision, however it is actually a very biased data set, and used solely here to show how DCA works, when it works (IE should the price of later serial payments have decreased) if instead the value of the investment has increased then DCA is inferior to lump sum contributions..
Your point is correct, but over the medium-term, I would say that the chart is a good approximation for how various stock market sectors behave. The stock market moves in cycles - secular/cyclical bull/bear markets. For 5 years, it might be moving mostly up (i.e, 2009-2014), but it might drop 50% next year. That's where DCA comes into play - broad sectors and over 3+ year time horizons.

However, if one is picking a specific stock, then, just as you said - don't DCA. If you feel the need to DCA after carefully analyzing a single security and deciding to invest in it, then you already know you shouldn't be picking stocks (because you're not confident of your analysis) :)
 

Matt

Administrator
Staff member
Your point is correct, but over the medium-term, I would say that the chart is a good approximation for how various stock market sectors behave. The stock market moves in cycles - secular/cyclical bull/bear markets. For 5 years, it might be moving mostly up (i.e, 2009-2014), but it might drop 50% next year. That's where DCA comes into play - broad sectors and over 3+ year time horizons.

However, if one is picking a specific stock, then, just as you said - don't DCA. If you feel the need to DCA after carefully analyzing a single security and deciding to invest in it, then you already know you shouldn't be picking stocks (because you're not confident of your analysis) :)
Very true, but what happens if this cycle is 10 years?
 

SC Parent

Level 2 Member
Very true, but what happens if this cycle is 10 years?
Do you mean if the bull market lasts 10 years? I don't see that affecting the math or the cost-benefit analysis. Say it takes you 3 years to DCA into your preferred allocation - you still have 7 years of the bull market to enjoy. Not as good as going all-in, but not terrible either. Conversely, if it's a 10-yr bear market you've missed some of the early losses, even though you have another 7 years of losses before your investments would start to recover. Not as bad as having gone all-in, but not terrible either. I see DCA as a volatility-reducing technique similar to regular portfolio rebalancing (which is DCA by another name, or vice versa).

I recognize that my own investment philosophy biases my views, as I'm not a believer in ever being "fully invested." And, I'm talking about broad market sector exposures here, not individual stocks. Most investors DCA into the market when they put their $50/month or $5,000/month into their 401(k) or brokerage accounts.

If one has $10,000 they're investing, does it matter if they DCA? I don't think it really matters. If they're in their 20s or 30s? No again. But, as one's investing time horizon shortens or as the amounts get larger, DCA is a way to avoid getting too much on the wrong side of a bet, especially resulting from exogenous shocks or (*this could never happen*) an incorrect analysis. (Of course, as one approaches the limits from the other side (very high net worth or a very short investing horizon, perhaps due to age/illness), then DCA again doesn't matter.)

DCA for portfolio allocation doesn't need to be a drip-feed, but it can be 2-5 trades rather than one. Transaction costs are negligible nowadays. I think the time horizon for the Vanguard table was too short and the sectors too broad. What that chart tells me is that after a year markets are up twice as often as they're down. It only looks at stocks and bonds, rather than more nuanced asset classes (though that's Vanguard's model). It's also over an 85 year investing horizon, which I don't have :) I'd be curious to see an analysis of 1965-1985 looking at small caps, large caps, international, precious metals, etc.

I'm not saying DCA is the best thing ever. Just that it's an investing tool that can be useful in maintaining discipline, reducing volatility or anxiety, and, in certain market situations, actually improving returns (though it can lower returns in other market situations).
 

Matt

Administrator
Staff member
Yep, I used to be a big fan of it, but the more I consider it the less in favor I am. While you certainly can become more 'average' with your basis, there is no way to know which direction the market is heading. As such, DCA really only offers any value if the asset declines during the drip feeding process.

If the asset appreciates during the process then you are both losing out on gains and increasing your basis when you buy in at the higher price.

Furthermore, once you are 'all in' you are 'all in' whether that be in a day, or over a few years. At that point in time you are at the mercy of the markets. I like your point on the $10K - it is where I am going with the "repair ratio" in that you need to have enough uncorrelated assets to protect the investment. Essentially with DCA you are allocating a shifting portion of one asset (cash) to another (the target asset) over time and providing the repaid ratio with that, but when the cash part has moved over into a fully vested DCA shift - you lose that power.

As such, my thoughts (which I am still shaping and about to put out in a post) is that rather than move from (EG) Cash>Stock in periodic investments, instead go all in with a well structured mix of less correlated investments, simply so that you aren't losing out on gains on the original cash but you also aren't exposed to a bear market without the ability to execute a repair to the asset.
 
Top