3 ms·

Is it possible to do it for dollar cost averaging? That is what happens if one didn't invest in a lumpsum but invested $1000 every month -- how does the returns

Is it possible to do it for dollar cost averaging? That is what happens if one didn't invest in a lumpsum but invested $1000 every month -- how does the returns look like for different time periods?

aeling 10y ago
It's discussed near the bottom of the writeup, but the author suggests somebody else may be better suited to do it: What about Dollar Cost Averaging (DCA) instead of Lump Sum Investing (LSI)? This one is going to be a little more difficult. I've included the original code and it's open-source, so if you want to play with the numbers, you can try to do this yourself. However, there are a couple points I'd like to keep in mind: •You're working with a "cash multiplier" as the main result. You'd be trading in a unitless scale for a scale with units since you'll need to specify a base value. Perhaps the act of buying a single share of SP500 per year might help with keeping the scale nondimensional? •There's inflation data in there, but at some point you're going to have to account for that if you're using dollars as a scale. How does one compare investing a dollar in 1902 vs. 2002? •DCA will probably cause some scaling issues. You'll need to accurately tune your algorithm to reflect an equivalent investment under the LSI algorithm I provide. My prediction is that DCA will "thin out" the plot and bring a lot more values closer to the average. This makes investing a bit safer, but at the cost of slower gains.
prattmic 10y ago
It is only over 10 years, but Vanguard has a short analysis of Dollar Cost Averaging at https://pressroom.vanguard.com/nonindexed/7.23.2012_Dollar-cost_Averaging.pdf https://pressroom.vanguard.com/nonindexed/7.23.2012_Dollar-c...
- pilom 10y ago
  The gist of their analysis was "since the market always goes up eventually, investing all up front will give you higher average returns but significantly higher variability"

Is it possible to do it for dollar cost averaging? That is what happens if one didn't invest in a lumpsum but invested $1000 every month -- how does the returns look like for different time periods?

aeling 10y ago

It's discussed near the bottom of the writeup, but the author suggests somebody else may be better suited to do it: What about Dollar Cost Averaging (DCA) instead of Lump Sum Investing (LSI)? This one is going to be a little more difficult. I've included the original code and it's open-source, so if you want to play with the numbers, you can try to do this yourself. However, there are a couple points I'd like to keep in mind: •You're working with a "cash multiplier" as the main result. You'd be trading in a unitless scale for a scale with units since you'll need to specify a base value. Perhaps the act of buying a single share of SP500 per year might help with keeping the scale nondimensional? •There's inflation data in there, but at some point you're going to have to account for that if you're using dollars as a scale. How does one compare investing a dollar in 1902 vs. 2002? •DCA will probably cause some scaling issues. You'll need to accurately tune your algorithm to reflect an equivalent investment under the LSI algorithm I provide. My prediction is that DCA will "thin out" the plot and bring a lot more values closer to the average. This makes investing a bit safer, but at the cost of slower gains.

prattmic 10y ago

It is only over 10 years, but Vanguard has a short analysis of Dollar Cost Averaging at https://pressroom.vanguard.com/nonindexed/7.23.2012_Dollar-cost_Averaging.pdf https://pressroom.vanguard.com/nonindexed/7.23.2012_Dollar-c...

pilom 10y ago
The gist of their analysis was "since the market always goes up eventually, investing all up front will give you higher average returns but significantly higher variability"