I have a script I like to run that receives as input my retirement-contribution history - amounts and dates. It pretends I simply bought VFINX every time (using adjusted-close history to account for dividends) and tells me the resultant APY. As of today, over a 20-year history, that would be 8% nominal. Factor in inflation over that time and it's below 6%. And if you were to graph that out, probably 90% of those monthly data points are much lower.
People in general aren't going to have the discipline to stay in the S&P-500 since it is so volatile, but any other approach is in effect trying to "beat the market" which doesn't work. I didn't stay in S&P-500 and my personal APY is lower.
The fact of the matter is, people tend to have more money to invest when times are good (and market is high), and less money to invest when times are bad (and market is low). So they're by definition not going to be able to match "historical stock market averages". So any advice from these many articles and blog posts basically come down to "in investment performance, be an outlier." Well, sure.
Median was 4.1% real, then that means a real return at 90% or 95% confidence was even lower.
So in my opinion, the graph confirms the rule of thumb.
10% inflation adjusted I'd think is pretty ridiculous, either it's short-term, not inflation adjusted or he's a one in a million financial miracle worker, but you can't not have any of them. Actually the S&P500 return since inception is about 10%, maybe he's referencing it. But the real return of 1950 to 2010 or so has averaged about 7%. And that uses CPI numbers which may underestimate price increases, particularly in some areas. (e.g. you probably want a better ROI than the national average in hotspots like San Fran considering you need to keep up with things like housing price increases if you're investing in stocks and living off of the returns).
But even there you need a strong stomach as these are just long-term figures in what in reality is a volatile market. It's pretty trivial to talk about a 'the real returns of 12% in the 1980s offset the 0% real returns of the 1970s' and conclude 'oh the average return ended up pretty good around 6%', but it's another thing to actually live through the 1970s, experience ridiculous inflation and stock returns that can't keep up, and sticking with your 'live off of capital' strategy, when nobody promised the 1980s would be different and you have a kid who's going to college by then, a spouse etc.
1. They usually cover about 100 years. If you're starting out your career you've probably got about 40-50 years of saving followed by say 20 years of retirement, i.e. your training data is not much more than the period you're trying to predict.
2. That history is primarily over the period of world history that can be summed up as "cheap oil". Personally I wouldn't use it to predict any other time period.
3. If you look at the returns over time, they're pretty much generational. Some generations get fantastic returns from the stock market, while others at best are keeping up with inflation. If you started investing in say 2000 you probably know which category you're in.
The 10 year US bond is often used as a risk free rate and it's about 2.2% right now if I remember correctly, and it's been as high as 3% in the past five years. And of course you can sell bonds too so it's not like they're illiquid, it's generally known as one of the most liquid markets on the planet.
But there's a bit more risk than than a 3 month bill of course which is also used as a risk free rate indicator quite a lot, and that's really low with inflation being so low right now.
Anyway I wouldn't call 4% incredible but it's definitely good and something many would be quite happy with.
Here's a nice little tool if you want to play with the numbers yourself over particular periods of time: http://www.moneychimp.com/features/market_cagr.htm
Over the 15-year period from 1966 to 1981, the real annualized rate of return in the stock market was -0.4% because of high inflation. You need to look at 20 to 30 year periods to get into the range of 7% real annualized returns across the 20th century. And there is the general issue with using the past as a statistical predictor of the future when the time window's length is a large fraction of the 20th century. Not to mention extreme outlier events whose level of impact isn't represented at all in the historical data.
A good book to read is Siegel's Stocks for the Long Run.
Granted, the 1950's had great returns, but the trend has been lower returns / year over time and both the 1970's and 2000's had negative real returns.
First, let's cover the tank in 10mm of it, which will give us all the protection we need. Freeing up that much room gives us more space to put a bigger gun on it. After vastly reducing the weight, even more so because we don't need to carry as much fuel, we don't need tracks, because an 8 wheel variant gives sufficient traction.
Oh, and since our new highly mobile and effective vehicle now costs a tenth as much, all the other expensive and slow vehicles taking on other roles in the army can be deprecated.
In fact, let's design a variant that carries fuel. In this way, instead of having 50 tanks advancing 100km per day, we can have 500 of our new vehicles advancing 500km per day, with the same firepower and effective armor, completely overwhelming any opposition. Why has nobody figured this out before?
OK, so, after we've taken the base assumption and followed the chain to its logical conclusion without error... we have nonsense. After realizing our mistake, and factoring in the real value for armor, our design and plans for reworking the army are absurd.
"Waiting longer" at 6% means 82 years to achieve the same returns as 50 years at 10%. But if we include 2% inflation as well, 6% takes 97 years to match 50 years at 10%. One of those is planning for your retirement if you're young... the other is planning for that of your grandchildren.
It also just so happened to quintuple his returns. Not exactly comparing apples to oranges when you have a 4% (realistic) v. 10% (unrealistic) return built into the model.
EDIT: I'm not sure why I'm getting downvoted here, he's talking about investing over a 20 year period - it's not unheard of to value an S&P index fund at 10% over that period!
The average annual inflation rate from 1970 is over 4%.
The annulaised return of the S&P 500 was 10.82% from 1970 to 1990. Which is grand but doesn't take into account inflation.
Adjust for inflation and the annualised return was 4.33%
So I am rolling to disbelieve that median 20 year return is 10%
It's also not unheard of for shares to double in value over a short period, but you don't know in advance which ones.
Quite simply, you are very unlikely to average at least 10% p.a. above inflation on investments over a long period.
The point is that eventually a quantitative difference becomes a qualitative difference. Few people can make a 10% annual return and many, many fewer still can make a 50% annual return. But if you just buy index funds, it's fairly straightforward for just about anyone to make a 4% annual (real) return.
Given the gigantic difference in how these curves grow over time, it's entirely fair to point this out.
It's like this popular meme:
http://i0.kym-cdn.com/photos/images/original/000/572/078/d6d...
Yes the beginning looks easy, yes I like where you got in the end, but you seem to have left out some important details on how you got there...
How can you get consistent 10% return on your investment?
Assuming a more realistic 6% nominal return - 2% inflation gives us a more realistic 4% real return.
Running the numbers that way ends up with a measly $10,000 in interest after 20 years not a whopping $47,000.