To clarify, and because it’s too late to edit my last post, the DJIA, which you are linking to, does not take into account normal dividends, only “unusual” dividends. The value of the index would be approximately 30 times higher if normal dividends were taken into account. E.g., this paper, written in 2000 which suggests a value 20 times higher, so I’m ad-hocing the last decade to arrive at a 30x result :).
OK, I realize that this is three posts in a row, and I realize that’s a bad thing, so this is my last word until someone responds:
You’re not necessarily taxed whenever you receive a dividend (e.g., a dividend may be a return of capital), the tax rates on types of dividends differ (e.g., qualified dividend v. common income), and there are tax strategies to avoid paying taxes on dividends regardless of their classification (e.g., hold stocks through Roth accounts). In addition, if you hold stocks you can tax loss harvest whenever you have a downturn, and you can roll those losses forward, so there are all sorts of ways to defer dividend taxes.
Given the complexity of assigning any tax rate to the analysis, and for purposes of comparing stocks to gold, I think it’s reasonable to treat dividends as non-taxable events (e.g., assume we’re holding the stocks in a tax-advantaged account, which is how the majority of average Joes are going to hold stock in any event). These tax-free accounts weren’t available going back to 1929, but I bet there were equivalent ways of avoiding dividend taxes.
In that case, it’s fair to compare the dividend-adjusted stock value multiplier to the gold value multiplier.
Let’s say the dividend adjustment is between 20 and 30 (say it’s 25); then the DJIA multiplier is approximately 1061x (rather than 42.5) and the gold multiplier is still 58. Stocks performed approximately 18 times better than gold over the single worst period you could pick for stocks (because we’re choosing 1929 as our starting date). Pick a different starting date and stocks look much much better.
Hm. You’re right. I was misled by the phrase, “Close price adjusted for dividends and splits” on the Yahoo page. I now assume that that column (Adjusted Close) was just adding the dividends valid at the time, without actually reinvesting the dividends back into the stocks. Reading the paper you referenced and looking at the numbers in the Yahoo table, it now seems obvious that they weren’t DRIPping back in.
Okay; everyone get back into equities!
[re reinvestment]
I’ve noticed that Sharebuilder reinvests all dividends from my stocks back into the individual stocks at no cost, whereas on Scottrade, my dividends just get added to my cash balance. The difference, I assume, is because Sharebuilder’s basic investments are made with fractional shares. I’d created accounts with both because Sharebuilder charges $4 per “scheduled” monthly investment, but $20 for individual trades, whereas Scottrade charges only $7. Now I’m thinking I should keep my longer term blue chips on Sharebuilder, with short term trades (or any trades in non-dividend paying stocks) on Scottrade.
[/reinvestment]
So, what accounts for your Evil username? It seems you’ve helped me figure some things out today.