Past performance vs future returns

[boggler]

Everyone says past performance is not indicative of future returns. I even
agree with it, largely because I believe in the EMH (at least to the extent
that I am not going to involve myself in high-frequency trading or
arbitrage). Yet this seems to run counter to a lot of other things in
life:

- When you decide to go work for a company because you think it will be
successful, isn't that just as much of a wager?
- When you buy a house because you think it's a "good price", isn't that
like buying a stock you think is undervalued?
- When you choose to write a book or start a company, thinking that you're
going to be successful, isn't this the same type of risk? (Many many
people start companies. Why will you be successful? Similarly, many many
people buy stocks thinking they are right and the other guy is wrong. Who
knows?)

I'm curious to know how others on this forum think about these issues in
relation to investing and the indexing approach.

[Nicho_1978]

I often wonder the same question myself. Maybe a better phrase is " past performance does not garauntee future return, but then again neither does the prospect of future earnings"

[Call_Me_Op]

You have misinterpreted and/or misquoted the statement. Past performance does not guarantee future returns. They are somewhat indicative - particularly over very long time periods - and particularly if you are looking at past returns of asset classes as opposed to individual securities or funds.

[hlfo718]

If past perf can predict future results, then Fidelity Magellan will be managing trillions of dollars since that was the most well known and one of the best performing fund from the 80's. Ditto for Legg Mason Value, CGM Cap Development, Berger 100, 20th Century Ultra, Kaufman... Vanguard will have a declining client and asset base since most people will simply buy last year's winners. If life was that easy brokers will go out of business. But off course that is the selling point of brokers, they can pick tomorrow's winners.

[EternalOptimist]

I think if you make any decision that's not based at least somewhat on past performance, you are foolish. Would you rather make a fund pick based on no information of its historical behavior Its just a caveat statement so they don't get sued.

[Kozmig]

I'm relatively new to these boards (at least posting), and am wondering what the thoughts of this board are on this question:

If you have a choice of 2 funds within the same asset class (could be large cap stocks funds, mid-cap, etc), and you had a choice between a Vanguard fund that is at a low expense ratio (say 0.06%) with 1 and 3 year returns of 15% and 8% respectively, or a managed fund with an expense ratio of 0.6%, but 1 and 3 year returns of 18% and 12% respectively, which would you select and why?

Reading through a lot of the responses on the portfolio reviews, I see a lot of people recommending the vanguard funds over the managed ones. Does this have to do with the question of this thread, i.e. not a guarantee of future results, is it an aversion to high expense ratios, or is it because when you just track an index in a vanguard fund, you are minimizing the variables, such as change of fund management?

[chaz]

Momentum sometimes helps.

[bertilak]

I'm relatively new to these boards (at least posting), and am wondering what the thoughts of this board are on this question:

If you have a choice of 2 funds within the same asset class (could be large cap stocks funds, mid-cap, etc), and you had a choice between a Vanguard fund that is at a low expense ratio (say 0.06%) with 1 and 3 year returns of 15% and 8% respectively, or a managed fund with an expense ratio of 0.6%, but 1 and 3 year returns of 18% and 12% respectively, which would you select and why?

Reading through a lot of the responses on the portfolio reviews, I see a lot of people recommending the vanguard funds over the managed ones. Does this have to do with the question of this thread, i.e. not a guarantee of future results, is it an aversion to high expense ratios, or is it because when you just track an index in a vanguard fund, you are minimizing the variables, such as change of fund management?

Go with the lower expense.

The higher returns may be because the managed funds happen to be taking on more risk. You can manage risk better by establishing an appropriate Asset Allocation. With the managed funds you can't be sure what level of risk you are taking on. Perhaps the higher historic returns are not actually enough to justify their (unknown?) risk level.

Another point: you have no way of determining if those returns will continue to outpace the index funds. Not enough information is available. Perhaps they were just lucky for three years. Perhaps the fund will change strategies after you buy in. Perhaps the "hot" part of the market that produced that excess return will turn cold.

[Kozmig]

I'm relatively new to these boards (at least posting), and am wondering what the thoughts of this board are on this question:

If you have a choice of 2 funds within the same asset class (could be large cap stocks funds, mid-cap, etc), and you had a choice between a Vanguard fund that is at a low expense ratio (say 0.06%) with 1 and 3 year returns of 15% and 8% respectively, or a managed fund with an expense ratio of 0.6%, but 1 and 3 year returns of 18% and 12% respectively, which would you select and why?

Reading through a lot of the responses on the portfolio reviews, I see a lot of people recommending the vanguard funds over the managed ones. Does this have to do with the question of this thread, i.e. not a guarantee of future results, is it an aversion to high expense ratios, or is it because when you just track an index in a vanguard fund, you are minimizing the variables, such as change of fund management?

Go with the lower expense.

The higher returns may be because the managed funds happen to be taking on more risk. You can manage risk better by establishing an appropriate Asset Allocation. With the managed funds you can't be sure what level of risk you are taking on. Perhaps the higher historic returns are not actually enough to justify their (unknown?) risk level.

Another point: you have no way of determining if those returns will continue to outpace the index funds. Not enough information is available. Perhaps they were just lucky for three years. Perhaps the fund will change strategies after you buy in. Perhaps the "hot" part of the market that produced that excess return will turn cold.

Thanks for the reply. It's always interesting to see what other people's thought processes are like.

[nydad]

I think it depends on the asset class. For some asset classes, the indexes may not be that great, and active management may do a better job over a long period of time. This wouldn't apply for US equities for example, but in foreign equities where indexes are newly developed and still evolving there may still be a role for active management. Many bond funds are actively managed as well, as another example, but some are more "active" than others...

[Jerilynn]

It's a paradox. If past hx doesn't tell us anything, then why do we buy ANY equities?

[midareff]

If past perf can predict future results, then Fidelity Magellan will be managing trillions of dollars since that was the most well known and one of the best performing fund from the 80's. Ditto for Legg Mason Value, CGM Cap Development, Berger 100, 20th Century Ultra, Kaufman... Vanguard will have a declining client and asset base since most people will simply buy last year's winners. If life was that easy brokers will go out of business. But off course that is the selling point of brokers, they can pick tomorrow's winners.

You are taking the previous statement and interpreting it incorrectly IMHO. Magellan, LMVTX, and other funds which performed well are subject to manager's errors in asset selection. That is a very different asset than a Total Market index, SV Index, International Index, etc. Take manager selection out of the process and you will be much closer to what Call_me was talking about. Also.. I did not see "predict" in his statement.

[midareff]

I think if you make any decision that's not based at least somewhat on past performance, you are foolish. Would you rather make a fund pick based on no information of its historical behavior Its just a caveat statement so they don't get sued.

Let's take an example of a future fund pick..... let's say it is now July and VG has opened the International Bond Index to investors. Since the fund has no prior history are all of those investing in it, and all of those holding other funds that will hold that fund, FOOLISH? By your statement they are.

[bertilak]

I think if you make any decision that's not based at least somewhat on past performance, you are foolish. Would you rather make a fund pick based on no information of its historical behavior Its just a caveat statement so they don't get sued.

Let's take an example of a future fund pick..... let's say it is now July and VG has opened the International Bond Index to investors. Since the fund has no prior history are all of those investing in it, and all of those holding other funds that will hold that fund, FOOLISH? By your statement they are.

Well, Vanguard as a whole has some history to go by. In any case, I would still hold off (for a while) on investing in a new fund in case there are any wrinkles that need to be worked out.

But I think the statements like "The performance data shown represent past performance, which is not a guarantee of future results." (from a Vanguard web page) are right on. There is no guarantee.

Past performance in the aggregate does tell us a great deal. For example, it tells us that good past performance of a particular stock, fund or manager is not likely to be extrapolated into the future. Index funds are a bit different. We don't expect a "good" run to continue indefinitely. Instead we expect the fund to track the index. This is more predictable and past performance can come into play as a predictor.

[midareff]

I'm relatively new to these boards (at least posting), and am wondering what the thoughts of this board are on this question:

If you have a choice of 2 funds within the same asset class (could be large cap stocks funds, mid-cap, etc), and you had a choice between a Vanguard fund that is at a low expense ratio (say 0.06%) with 1 and 3 year returns of 15% and 8% respectively, or a managed fund with an expense ratio of 0.6%, but 1 and 3 year returns of 18% and 12% respectively, which would you select and why?

Reading through a lot of the responses on the portfolio reviews, I see a lot of people recommending the vanguard funds over the managed ones. Does this have to do with the question of this thread, i.e. not a guarantee of future results, is it an aversion to high expense ratios, or is it because when you just track an index in a vanguard fund, you are minimizing the variables, such as change of fund management?

Go with the lower expense.

The higher returns may be because the managed funds happen to be taking on more risk. You can manage risk better by establishing an appropriate Asset Allocation. With the managed funds you can't be sure what level of risk you are taking on. Perhaps the higher historic returns are not actually enough to justify their (unknown?) risk level.

Another point: you have no way of determining if those returns will continue to outpace the index funds. Not enough information is available. Perhaps they were just lucky for three years. Perhaps the fund will change strategies after you buy in. Perhaps the "hot" part of the market that produced that excess return will turn cold.

Thanks for the reply. It's always interesting to see what other people's thought processes are like.

It has ben well documented in academic research that the costs of active management; the research, trading costs, high payroll, etc., and the higher fees make it extremely difficult if not impossible for an actively managed fund to perform better than is asset class index cousin. From memory, 80% of the actively managed funds underperformed their benchmark handily at the five year mark. I can't speak for you but I would rather be in the 20% that did not underperform their benchmark (a few basis points of fees excluded). Try to remember that most funds are operated for the benefit of the owner, and most are not owned by the investors.

[investor]

Max Heyne and Michale Price did one heck of a good job with Mutual Shares from 1949 till when Price sold it all in the mid 1990's. Was quite a track record for active management...very lucky I guess. Value investing at its finest.

investor

[midareff]

I think if you make any decision that's not based at least somewhat on past performance, you are foolish. Would you rather make a fund pick based on no information of its historical behavior Its just a caveat statement so they don't get sued.

Let's take an example of a future fund pick..... let's say it is now July and VG has opened the International Bond Index to investors. Since the fund has no prior history are all of those investing in it, and all of those holding other funds that will hold that fund, FOOLISH? By your statement they are.

Well, Vanguard as a whole has some history to go by. In any case, I would still hold off (for a while) on investing in a new fund in case there are any wrinkles that need to be worked out.

While it may not make sense to be the very first in line all of that sort of missed my point. This is a fund with no history, which will be an index fund to an index of an asset class that does have a history.

"According to Vanguard, the new bond fund will track the Barclays Global Aggregate ex-USD Float Adjusted RIC Capped Index (USD Hedged). The index comprises approximately 7,000 high-quality corporate and government bonds (average credit quality AA2/AA3) from 52 countries."

[EternalOptimist]

I think if you make any decision that's not based at least somewhat on past performance, you are foolish. Would you rather make a fund pick based on no information of its historical behavior Its just a caveat statement so they don't get sued.

Let's take an example of a future fund pick..... let's say it is now July and VG has opened the International Bond Index to investors. Since the fund has no prior history are all of those investing in it, and all of those holding other funds that will hold that fund, FOOLISH? By your statement they are.

If the info is available and you chose to ignore is what I meant... the OP is talking about past performance vs future returns.

[bertilak]

While it may not make sense to be the very first in line all of that sort of missed my point. This is a fund with no history, which will be an index fund to an index of an asset class that does have a history.

Right. I realized that myself and was updating my post at the same time you posted. Look at it now.

[midareff]

While it may not make sense to be the very first in line all of that sort of missed my point. This is a fund with no history, which will be an index fund to an index of an asset class that does have a history.

Right. I realized that myself and was updating my post at the same time you posted. Look at it now.

Exactly......

[LadyGeek]

This thread is now in the Investing - Theory, News & General forum (investing).

[Lumpr]

Everyone says past performance is not indicative of future returns. I even
agree with it, largely because I believe in the EMH (at least to the extent
that I am not going to involve myself in high-frequency trading or
arbitrage). Yet this seems to run counter to a lot of other things in
life:

- When you decide to go work for a company because you think it will be
successful, isn't that just as much of a wager?
- When you buy a house because you think it's a "good price", isn't that
like buying a stock you think is undervalued?
- When you choose to write a book or start a company, thinking that you're
going to be successful, isn't this the same type of risk? (Many many
people start companies. Why will you be successful? Similarly, many many
people buy stocks thinking they are right and the other guy is wrong. Who
knows?)

I'm curious to know how others on this forum think about these issues in
relation to investing and the indexing approach.

IMHO - Yes. At a generalized level, these decisions are all the same. However, there is an important difference in application. You have the option of diversifying risks related to your financial assets. Practically impossible to adequately diversify risks related to those other decisions.

[boggler]

Everyone says past performance is not indicative of future returns. I even
agree with it, largely because I believe in the EMH (at least to the extent
that I am not going to involve myself in high-frequency trading or
arbitrage). Yet this seems to run counter to a lot of other things in
life:

- When you decide to go work for a company because you think it will be
successful, isn't that just as much of a wager?
- When you buy a house because you think it's a "good price", isn't that
like buying a stock you think is undervalued?
- When you choose to write a book or start a company, thinking that you're
going to be successful, isn't this the same type of risk? (Many many
people start companies. Why will you be successful? Similarly, many many
people buy stocks thinking they are right and the other guy is wrong. Who
knows?)

I'm curious to know how others on this forum think about these issues in
relation to investing and the indexing approach.

IMHO - Yes. At a generalized level, these decisions are all the same. However, there is an important difference in application. You have the option of diversifying risks related to your financial assets. Practically impossible to adequately diversify risks related to those other decisions.

So why not invest in, say, Fidelity Contrafund or Berkshire Hathaway instead of index funds, or even investing in the stock market as a whole because it has done well in the past? How is this different from deciding to work at Google because you think it's a good company? Why do we balk at the former but accept the latter?

[larryswedroe]

Here's the thing
Think tennis, or chess, small difference in skill can lead to large differences in outcomes. Reason, competition is one on one.
With investing huge differences in skill can lead to no difference in performance because you are not competing against other individuals who you might have more skill than, but the collective wisdom of the market, which is much more difficult competitor. That means the nature of the competition is very different and thus while you can expect past performance to repeat in such one on one competitions it doesn't mean you should in a very different type of competition

The other reason is that success as say tennis player doesn't create more hurdles but success as active manager does, in fact success sows the seeds of its own destruction because more money flows in and that leads to higher hurdles (either market impact costs or closet index costs)

Hope that helps
Larry

[Akiva]

I'm relatively new to these boards (at least posting), and am wondering what the thoughts of this board are on this question:

If you have a choice of 2 funds within the same asset class (could be large cap stocks funds, mid-cap, etc), and you had a choice between a Vanguard fund that is at a low expense ratio (say 0.06%) with 1 and 3 year returns of 15% and 8% respectively, or a managed fund with an expense ratio of 0.6%, but 1 and 3 year returns of 18% and 12% respectively, which would you select and why?

Reading through a lot of the responses on the portfolio reviews, I see a lot of people recommending the vanguard funds over the managed ones. Does this have to do with the question of this thread, i.e. not a guarantee of future results, is it an aversion to high expense ratios, or is it because when you just track an index in a vanguard fund, you are minimizing the variables, such as change of fund management?

Conceptually if you break out the returns from that hypothetical active fund, you are going to see that some of it is intelligent risk taking (which the vanguard fund gives you), some of it is dumb luck, and some of it is skill. Now in the real world we can't distinguish between luck and skill. For one thing given reasonable assumptions about variances, etc. the amount of data we'd need to distinguish skill from luck is larger then you are ever going to get given the size of typical out-performance and the normal tenure of a fund manager. Second, we have enough data on actively managed funds that we know that shorter term returns are non-predictive of future returns on average. (And so we expect there to be regression to the mean, although most people don't understand what this means in practice.)

Consequently, we assume that all of the unaccounted for excess performance is luck. Now we compare what we expect our funds to return in the future (and luck has zero expected value). The Vanguard fund will return the market less .06% and the active fund will return the market less .6%. So we assume the Vanguard fund will do better.

[Akiva]

It's a paradox. If past hx doesn't tell us anything, then why do we buy ANY equities?

Well we have some theoretical considerations to go on that tell us how equities should be valued in abstract and that lead us to conclude that the historical out-performance of equities is not random error. (Though reasonable forward going expected out-performance is certainly lower than US history would indicate.)

[RobG]

There is another angle to the meaning of the phrase that nobody has mentioned. It is very important to know the reason for good returns and if they are indicative of what you can expect for the future. An example would be a long term government bond fund like Vanguard's EDV which is a long term STRIPS fund. To an inexperienced investor this fund appears to have a lot going for it:
1) it is a low expense Vanguard fund
2) It is U.S. bond so and we all "know" that means it is basically 100% safe
3) it has a proven track record (10% return since inception).

Now most of us know long term STRIPS went way up because bond rates went down. However, it is unrealistic to expect this in the future.

The recent run-up in gold is another example of past return not being indicative of its future likely return.

[YDNAL]

Everyone says past performance is not indicative of future returns.

Past performance "does not guarantee future returns."

For instance, you may hear that for 30 years (through 2009) the average annual return of the S&P 500 was 8.3%. HERE is data.

• Is that guaranteed for 2010-2030s ?... certainly not.
• Is that EVEN "indicative" ?... perhaps, perhaps not.

Life is full of uncertainties (except death and taxes, I hear)... so are investing/returns.

[dbr]

There are a lot of maxims and rules of thumb stated in investing, including in this forum.

I like to think of them as references to or titles for discussions that can be long and complicated, because some times certain things just aren't simple. Past performance and future returns is one of those long and complex discussions.

[Austintatious]

We're all making investment determinations based on our assessments of probabilities and, in doing so, we're relying heavily, though not exclusively, on past performance, even though we know it offers no guarantee. The three fund (even the two fund) portfolio investors are rolling the dice, based on their perceptions of the probability that, over the long haul, indexing is going to pay off. It's a pretty good bet, in my opinion, but it's still a bet and it's made largely because of past performance. If they didn't think it was reasonable to take that bet, they'd have all the savings in FDIC covered CD's, I-bonds and other bets they perceive to be safer. Even those placing their faith in the government to make sure their CD's and their savings accounts are insured are gambling that there will be no disruption of society so severe that even our governments promises might not be kept. If we'd experienced an event in our past that had caused the government to default on its promises to make good on insuring our deposits, we'd be thinking more about placing our wealth somewhere else. Probabilities, determined largely but not exclusively based on past performance, right?

[boggler]

Here's the thing
Think tennis, or chess, small difference in skill can lead to large differences in outcomes. Reason, competition is one on one.
With investing huge differences in skill can lead to no difference in performance because you are not competing against other individuals who you might have more skill than, but the collective wisdom of the market, which is much more difficult competitor. That means the nature of the competition is very different and thus while you can expect past performance to repeat in such one on one competitions it doesn't mean you should in a very different type of competition

Agreed. However, all of the examples I noted above: company, book, etc. are all instances of an individual making a decision against a market, so it seems similar in nature to the investing case. How are they any different?

[William4u]

If past performance was not a reliable indicator of future performance, then I would not be invested in the stock market at all.

[boggler]

If past performance was not a reliable indicator of future performance, then I would not be invested in the stock market at all.

Exactly. So where do we draw the line?

[William4u]

If past performance was not a reliable indicator of future performance, then I would not be invested in the stock market at all.

Exactly. So where do we draw the line?

In the past, certain types of past performance did not correlate to future performance (past active fund outperformance, for example, did not correlate with the "future" outperformance that is now in the past too).

[SSSS]

Fortunately for our civilization, we have a handy thing called "statistics" that can tell us whether a correlation exists.

It can be statistically proven that stocks have higher volatility and higher expected returns than bonds, for example.

It can be statistically proven that there is NO correlation between past outperformance and future outperformance of an active investment manager.

[Akiva]

Fortunately for our civilization, we have a handy thing called "statistics" that can tell us whether a correlation exists.

It can be statistically proven that stocks have higher volatility and higher expected returns than bonds, for example.

It can be statistically proven that there is NO correlation between past outperformance and future outperformance of an active investment manager.

Strictly speaking the statistics will only tell you that stocks have higher historical returns. The future could always be different because prices are determined by human action and people could act differently in the future. To infer about future expectations from past data, you have to make some assumptions that allow you to extrapolate. Thus you ultimately have to fall back on theory to justify the belief that stocks have higher expected return.

Furthermore, the consensus is that the history of the US stock market is abnormal and affected by numerous biases that when properly accounted for reduce the risk premium for stocks quite dramatically. Lots of people assume 5% but realistically it is probably more like 3%. And that 3% premium is only for index fund investors and other people who invest efficiently. There are plenty of studies looking at the accounts of investors who trade stocks themselves that show that the typical stock investor captures no risk premium at all...

[Taylor Larimore]

Boggler:

Everyone says past performance is not indicative of future returns.

Not "everyone," but the experts I quoted in this article agree with you:

Taylor's post on Past Performance

Best wishes.
Taylor