next »

[sjsmith]

Sjsmith, this is informative. When you say "not taking account of inflation", do you mean deflated or not deflated. If your numbers are real, i.e. deflated, they're pretty big. If they're nominal, i.e. not deflated, they're terribly small. [I should have a feel for which are the right numbers, but, sorry, I don't].

My not taking into account inflation I mean I just read the value of the Dow at the specific years I mentioned straight off the yahoo chart and that obviously doesn't take into account inflation.  So for example, the DJ in 1990 is about 2700.  In 1950 it is about 200.

(2700/200) = 13.5

(13.5)^(1/50)-1=6.7% average compounded return over those 40 years.

The variety in the numbers in "long-term" periods (any period over 10 years) really show that not all long-terms are created equal.  Timing is still vital.  You could invest for 30 years and easily get practically no growth or even negative real growth.  Or you could invest for 30 years and get massive growth.  It really depends on timing.   

I merely skimmed this http://www.ssab.gov/Publications/Financing/estimated%20rate%20of%20return.pdf. A long-term real annual return somewhere under 7% seems plausible [far broader than DJ, account for dividends, etc., etc.]. This compares to a real "safe" return of ca. 3 ppa on long term Treasuries. So, you are being paid 4 ppa for bearing the risk. We can all form an impression if this is worth it to us at various times in our lives [years until retirement, etc., etc.] if we consider volatility. I expect this information may help mdwlark, too.

Yes, but that 6.7% is for a specific period of 40 years.  That is my whole point.  Not all long-terms are created equal.  The history of the Dow shows you cannot just sink your money into stocks and blindly trust that in the "long-term" you will make good returns.  There doesn't seem to be a real cut-off point where you can say after X number of years you can expect your average rate of return to be around Y.  Which is my definition of an accurate long-term prediction of growth.  Timing is still pretty vital.  Or you need to be able to choose good specific stocks.   

To tell the truth I can't really detect a pattern in the changes of growth rate you can see in the Dow historical charts.  It does seem to be something chosen by the fundamentals of the time the change occurs.  So once again you cannot blind trust that you will get certain returns over the "long-run".  In fact when you think about it the ultimate long-term return from beginning to end has to be around zero.  Because assuming exponential growth any finite growth rate will result in the Dow shooting up to infinity.  Or exponential growth only applies to specific periods of the Dow (with constant slopes in the semi-log plot) and does not apply in the ultimate long term.  As I said I wouldn't want to try to fit a single exponential function to that graph. 

[dismalist]

I meant all of this descriptively, of the past. Who knows what the future will bring.

[pedanterast]

"There doesn't seem to be a real cut-off point where you can say after X number of years you can expect your average rate of return to be around Y."

I agree with this basic premise and that's why I think many of the early posters on this thread were overly sanguine.

However, Monte Carlo analysis is useful here.  However, a lot of the Monte Carlo calculators suffer from having assumptions buried in them that are mutually exclusive in reality.  For example, any given annual return assumption could be based on a very high rate of inflation and a very low short-term interest rate, and those two things would not normally happen at the same time.

The only thing an investor can eat at the end of the day (or the end of the career would perhaps be more apropos) is after-tax, inflation-adjusted total returns.  So, it's a waste of time to look at price-only returns such as the Dow.  Much better to look at the S & P total return index, which is (much) broader, includes dividends, and is capitalization weighted instead of price weighted.  If you take that, and adjust each year for inflation, and then look at both the geometric mean (compounded return) and the standard deviation, that produces decent input for a Monte Carlo simulation.  I did that back in the mid-90s, using the Ibbotsen data dating from 1926.  But, now we'd need to add 14 or 15 more years' worth of data, and I think that would result in a higher standard deviation and therefore worse results than what I got.  At that time I concluded that (having looked at multiple overlapping 20 year time horizons), somewhere around 5.5% compounded was a realistic estimate of a probable return on equity investments over a 20-year time horizon.

Now let me make a few comments.  First, why did I chose 20 year periods to study?  Well, I figured the average investor has 40 years to put money in (say ages 25 to 65, for example).  So, 20 years was the halfway point in that time span.  Second, once you reach the withdrawal stage, the pattern of returns is extremely important.  If the early years are bad and the end years are good, that does not produce the same result as when the early years are good and the end years are bad; it produces a worse result because the good returns get applied to an amount that has already been drawn down.

An investor looking at the "bad" 5 or 10 or 20 or 30 year periods would probably never invest in stocks.  An investor looking at the "good" periods would probably always invest in stocks.  I used a formula of 120 minus my age to determine the % of my portfolio to devote to stocks, and I was highly diversified the whole time, and I'm pretty disappointed right now.  I'm still 10 or 12 years from hitting the withdrawal stage, according to my original game plan.  Will those 10 or 12 years be good years?  My guess is no, with all the structural problems.  Now I have to decide what to do about that.  Here are some of the possibilities:

1)  Retire at age 160 instead of age 60.
2)  Reduce spending and increase investing.
3)  Change what I'm investing in which is currently (120 - my age) in equities, 5% in REITs, and the rest in fixed income.
4)  Increase my risk level.
5)  Drink more to reduce my life expectancy.

Probably some combination of 1, 2, and 5 looks best right now.

[dismalist]

Pedanterast, this is excellent. Advice #5 one should follow anyway, even if it doesn't reduce life expectancy. :-)

[mdwlark]

pedanterast, #3, what are you going to change?  Thanks! 

[pedanterast]

As I said, I am probably not going to change #3 although I might increase my international exposure somewhat.  I haven't decided yet.  My international exposure is (as of my last re-balancing; I re-balance quarterly) 20% of my total equity exposure.  I might go to 25 or 30%.  I have until the end of the year to decide that since I won't be re-balancing until then.  I did cheat a little bit and change my 401k contributions going forward to 100% TIAA-CREF traditional annuity, partly because I have very little Social Security built up.

Right now I am concentrating on cutting expenses and generating additional income (#2).  Of course this is inconsistent with #5 but what I am doing there is buying cheaper booze.  I figure if I buy 20% cheaper brands I can drink 10-15% more and still come out ahead.  Fortunately I can get a liter of somewhat drinkable wine for about $1.20 where I live.  My favorite beer was on sale 3 for 2 last week so I laid in a good supply of that as well.

[johnstevenson]

If you can't afford to lose it then don't do it.

[mdwlark]

If you can't afford to lose it then don't do it.

I can't say for everyone else, but I have already lost what should have been relatively safe investments (previous posts).  I need to rebuild a retirement in a relatively short period of time.  Doing nothing with what I have left means I won't have an adequate retirement.  Trying to build a retirement means risking what little I have.  Slogans just don't cover the dilemma.  I think a lot of people are facing tough decisions.  Have we hit bottom or are we going into Great Depression II?  Everything is a risk.   

[conjugate]

Right now I am concentrating on cutting expenses and generating additional income (#2).  Of course this is inconsistent with #5 but what I am doing there is buying cheaper booze.  I figure if I buy 20% cheaper brands I can drink 10-15% more and still come out ahead.  Fortunately I can get a liter of somewhat drinkable wine for about $1.20 where I live.  My favorite beer was on sale 3 for 2 last week so I laid in a good supply of that as well.

My problem is that I am fond of red wine, which may work to increase my life expectancy, so that I will need more money amassed for retirement.  To counteract this, I have asked my favorite liquor store if they have any cheap gin produced in lead-lined bathtubs, but they just look at me as if I were out of my mind.  I usually leave before they call the cops.  :-)

[dismalist]

Right now I am concentrating on cutting expenses and generating additional income (#2).  Of course this is inconsistent with #5 but what I am doing there is buying cheaper booze.  I figure if I buy 20% cheaper brands I can drink 10-15% more and still come out ahead.  Fortunately I can get a liter of somewhat drinkable wine for about $1.20 where I live.  My favorite beer was on sale 3 for 2 last week so I laid in a good supply of that as well.

My problem is that I am fond of red wine, which may work to increase my life expectancy, so that I will need more money amassed for retirement.  To counteract this, I have asked my favorite liquor store if they have any cheap gin produced in lead-lined bathtubs, but they just look at me as if I were out of my mind.  I usually leave before they call the cops.  :-)

Luckily, the relationship between wine consumption and life expectancy is not monotonic. If you increase wine consumption sufficiently, you will reduce your life expectancy. We wine drinkers can eat our cake and have it, too, so to speak.

[mdwlark]

At least taking up wine drinking would reduce the anxiety.

[dismalist]

At least taking up wine drinking would reduce the anxiety.

Yes!

On the more serious issue, you must decide how much risk you can bear. The only contribution I can make is to remind everybody that if consumption has to be cut, it should also be cut during retirement.