[roenxi]

> Historically, gold isn't that great, but that's past performance.

Crikey. What performance counts as good in your eyes? Gold has been a pretty rewarding investment for the last 20 years. I think the median annual performance for the last 20 years has been something like 10% in USD (worthless metric for an investor I know, but the point is that most years are very solid up years). That is a pretty reasonable performer in my low-risk book. 2010-2014 represented an unusual bad time to buy gold in a decade long trend of great years to buy gold.

In terms of actual return it depends when you pick your start and end date; but the fact that we sometimes have major crisis was always totally predictable and the major governments responses have been reasonably predicable. The only surprise at the moment is the specific fact that it is a pandemic and in 2020. The price pattern on the charts is not surprising.

[Retric]

It’s only over short time periods that gold might look like a good investment. Inflation adjusted gold is flat over the last 40 years. That’s a real ROI of 0 from 1980 to 2020. Which means for every good investment someone else lost money.

Now it’s currently above the historic average so you do get 1% real returns from 1900 to 2020. But, stop in say 2010 and things look even worse.

[jkhdigital]

This is the best chart I've encountered for such comparisons: https://www.longtermtrends.net/stocks-vs-gold-comparison/

There are extended periods when gold absolutely crushes everything, and periods when it is totally flat (or worse). In the long run, yes, it is a poor investment because it is an inert metal. But for certain periods of time it is absolutely a great alternative to bonds and equities.

[Retric]

Look at the bottom chart, gold got crushed in that time period when you include dividends. The top chart is effectively meaningless by excluding them as you can’t invest in the S&P 500 without gaining dividends.

As to gold having positive returns in some window, sure but if you can time the market you can make money investing in anything. It’s doing the timing that’s difficult not looking for windows in historic data.

[gridlockd]

> In terms of actual return it depends when you pick your start and end date

Exactly. The 20 years range is cherrypicked for to make gold look good, it doesn't generalize.

If you had bought stock (and reinvested dividends) at the top of the market in 2008, you would've been better off with stock today. If you had bought gold at the top in 2011, you would've been better off with stock today. If you had bought gold at the bottom in 2008, you would've been better off with stock today.

Really, the only times where gold would've outperformed (to date) in the last 50+ years is around the time of the dotcom bubble and the last ~2 years.

Realistically though, you're going to buy the market at its highs and lows, you're going to exit at some point when you need to, so it's highly subjective.

[refurb]

Gold has been a pretty rewarding investment for the last 20 years.

Sure. And a horrible investment over the last 40 years. You’d still have a negative return if you bought in 1980.

[roenxi]

Peering 40 years into the past to inform investment decisions is fine as far as it goes, but I feel that the investment environment today is different from what it was in 1980.

Gold doesn't do much over a century, it's main property of interest is that it sits in an inert fashion wherever it is left. I'm not too surprised on worried if it has a near-0 return over 50 years.

I don't actually believe it has a positive return over the last 20 years either; but most people don't accept my "the inflation rate is meaningless, stop referring to it to value real returns" argument so I usually don't bother mentioning it.