Author Topic: Another pro for dollar cost averaging (or investing on a schedule)  (Read 512 times)

BobTheBuilder

  • Stubble
  • **
  • Posts: 136
I just want to share thought that crossed my mind since Tesla dropped 12% yesterday and will probaly swing up again as violently in the near future: DCA
You all know the topic, I just made some computations based on an example and a real world case.

Say you invest 150$ a month in a specific security that has a high volatility. In a very simple model, that could mean that a stock oscillates like a sinusoidal wave between 5$ and 15$ monthly, and the mean value is 10$. Now if you buy two times at the mean price of 10$ a share, you end up with 30 shares after two months. If you hit the top and the bottom of the price, you end up with 150$ / 5 $ = 30 shares, plus 150$ / 15$ = 10 shares, which would be 40 shares total.

Of course this example is really simplistic, but it shows the concept.

Now I have spread-sheeted this one out for a ridiculously volatile stock I very much like. TESLA.
Over the last 6 months, I computed the mean share price per day as well as for all 6 months, and assumed buying shares for 1000$ per trading day (obviously only to keep the numbers simple), or 125,000 in a lump sum.

If you buy shares for the mean price over 6 months, you end up with 449.74 shares
If you buy shares for the mean price on each specific day, you end up with 454.87 shares

So in this example with a volatile single stock, you probably receive an upside of 1.13% for not caring about the news and not attempting to time the market.

This difference is so large, that it makes up the higher trading costs for smaller investments, which are in my specific case for an investment of 2000 9.9 for a single purchase (or 0.5%) or 1.5% if you make monthly contributions of a fixed sum. This example still ignores if a stock has an uptrend or downtrend.

Long story short: If you want to buy into specific stocks for fun, you might want to consider DCA if the target sum is comparably small (which it should be, because of the single equity risk). This also means that making your monthly contributions as your pay checks arrives instead of forking out a larger amount every few months after first accumulating said sum not only gives you more time in the market, but also possibly a minor volatility bonus. And peace of mind, because it's automated.




MustacheAndaHalf

  • Handlebar Stache
  • *****
  • Posts: 1688
Re: Another pro for dollar cost averaging (or investing on a schedule)
« Reply #1 on: January 19, 2019, 09:19:00 AM »
Don't forget costs matter.  If you invest $75 twice a month, and pay a $7 commission, you're losing nearly 10% of your investment immediately.  Buying $150 worth once a month cuts that $7 commission down to ~5%.  Regular investment is important, but also consider what you're paying to make that investment.

Instead of buying an individual stock, you could buy a diversified mutual fund for $0/trade.  At Vanguard, most ETFs (~1800+) are $0/trade.  Those seem like better approaches than buying into single companies and paying a commission.

flipboard

  • Stubble
  • **
  • Posts: 147
Re: Another pro for dollar cost averaging (or investing on a schedule)
« Reply #2 on: January 19, 2019, 09:34:53 AM »
If you think DCA is nice... you should look into Value Averaging.Provides even more benefit over DCA, at only little more complexity. It's easiest just to read "Value Averaging" by Edleson. (I'm still reading it myself...)

$7 commissions certainly aren't realistic though, considering there are brokers charging either $0 or $1 per trade. I would however mirror the suggestion of using DCA or VA for diversified holdings as opposed to single stocks.

BobTheBuilder

  • Stubble
  • **
  • Posts: 136
Re: Another pro for dollar cost averaging (or investing on a schedule)
« Reply #3 on: January 19, 2019, 10:02:54 AM »
MustacheAndaHalf you are right, of course the costs matter. In this specific example, my broker charges 1.5% for a monthly plan buying a single stock, but no fixed fee. E.g. buying 6 times costs the same as buying 3 times: 30 for a 2000 single stock investment instead of 9.9 for buying only once. The whole point is that DCA is more expensive in this case, but the benefit of buying low sometimes and buying high sometimes vs buying a (random) mean value outweighs the increased transaction costs. For an ETF investment, 1.5% would be stupid expensive

There are of course ETFs available with no buying fee at all. On the other hand, individual stocks have an expense ratio of 0. If you hold an individual stock forever, you have no recurring cost.

I am not advocating buying single stocks instead of ETFs. ETFs are more cost effective and give broad diversification. This is only for the case of buying single stocks in limited investments as a small part of a portfolio that should be constructed of ETFs. I would not want to try managing 1000 stocks of 1000$ each if I could just buy an index instead.
Value averaging is an interesting concept for volatile market phases too!