Desi;35289 So the risk of a loss in much higher in a shorter term investment than in a longer term in the equities market. [/quote]
I hear the above statement often, and do not buy it. In what sense is it a valid statement? How is risk defined in the statement above?
Desi;35289 I will address why tomorrow unless someone else addresses prior to that. [/quote]
OK. Will pursue the issue after you address it. Others who subscribe to the statement above may also feel free to defend / elaborate.
Is equity investment safer over long term ?
Is equity investment safer over long term ?
Is equity investment safer over long term ?
Bobus;35295I hear the above statement often, and do not buy it. In what sense is it a valid statement? How is risk defined in the statement above?
OK. Will pursue the issue after you address it. Others who subscribe to the statement above may also feel free to defend / elaborate.
So the risk of a loss in much higher in a shorter term investment than in a longer term in the equities market.
This is a controversial issue in finance and there have been many research articles supporting either side of the issue. The essence of the argument is this:
Holding stocks over long time periods reduces the standard deviation of real returns and also reduces the chances of the portfolio failing to earn a real return. Historically, if you look at the real returns for holding periods greater than 17 years or more stocks never offerred investors a negative real return. As holding period increases the possibiity of negative real return reduces - at least for US stocks from 1871 onwards. This gave rise to the idea that stocks are less risky in the long run than in the short run.
Arguing against this are finance heavy weights, Merton, Samuelson and Bodie who hold that longer holding periods imply higher not decreased risk. They make a pretty strong case from first principles and option based arguments that longer holding periods do imply higher risk.
I have developed a few grey hairs trying to figure both arguments out. Here is a simplified description of how I think about this:
If you assume that stock markets do not get wiped out completely and further assume that economic growth more or less progresses forward, then the stock markets are less risky for long time horizons in the sense that investors have a higher chance of earning positive real returns or less chance of earning negative returns longer the investment horizon. Does that mean Merton, et all are wrong? No. Even though you might have lower risk as above, there is a higher risk of absolute $ variation. Say you invest $100K, then the risk that your investment would be less than $X K (i.e. some positive real return) is reduced the longer your investment horizon but the absolute deviation can be quite large i.e. it could be $1000K or $250K as time period increases. Most people experience/feel risk in the manner described first, even though Merton, et all is the more accurate theoretical definition of risk.
Vinod
Is equity investment safer over long term ?
Vinod: Thanks for the response.
Holding stocks over long time periods reduces the standard deviation of real returns and also reduces the chances of the portfolio failing to earn a real return.
Just to make sure, by real, you mean nominal return less inflation?
Do you mean standard deviation of annualized return or of total return?
Say you invest $100K, then the risk that your investment would be less than $X K (i.e. some positive real return) is reduced the longer your investment horizon but the absolute deviation can be quite large i.e. it could be $1000K or $250K as time period increases.
Are you saying that the long term investment stochastically dominates (in the sense of first order stochastic dominance) the short-term investment? What has been highlighted in bold above suggests so.
http://en.wikipedia.org/wiki/Stochastic_dominance
Also, what measure of risk do Merton et al use?
Holding stocks over long time periods reduces the standard deviation of real returns and also reduces the chances of the portfolio failing to earn a real return.
Just to make sure, by real, you mean nominal return less inflation?
Do you mean standard deviation of annualized return or of total return?
Say you invest $100K, then the risk that your investment would be less than $X K (i.e. some positive real return) is reduced the longer your investment horizon but the absolute deviation can be quite large i.e. it could be $1000K or $250K as time period increases.
Are you saying that the long term investment stochastically dominates (in the sense of first order stochastic dominance) the short-term investment? What has been highlighted in bold above suggests so.
http://en.wikipedia.org/wiki/Stochastic_dominance
Also, what measure of risk do Merton et al use?
Is equity investment safer over long term ?
Bobus,
Just to make sure, by real, you mean nominal return less inflation?
Do you mean standard deviation of annualized return or of total return?
Yes I mean real as nominal less inflation and the std dev is for the annualized return i.e. geometric return. Basically for stocks the std dev of geometric return falls as time period increases. (the main idea of Jeremey Siegel's Stock for the long run, where as it increases for bonds)
Are you saying that the long term investment stochastically dominates (in the sense of first order stochastic dominance) the short-term investment? What has been highlighted in bold above suggests so.
No. Note the big assumption regarding continuity of the stock market i.e. we have assumed away a big risk of the stock market that it does continue to operate and economic growth continues. This might seem silly but many stock markets have indeed been wiped out in the last 100 years. This kind of risk does increase when you increase the time horizon. Think of it this way, is the probability of a bad economic event more likely in the next 1 hour or more likely over the next 100 years? Second assumption is regarding how we think of risk. The assumption is we think of risk in terms of portfolio falling below some x level rather than in terms of variance of the total portfolio which is a more accurate way of looking at risk but that is not how most people perceieve of risk.
Also, what measure of risk do Merton et al use?
Std dev of total return.
Vinod
Just to make sure, by real, you mean nominal return less inflation?
Do you mean standard deviation of annualized return or of total return?
Yes I mean real as nominal less inflation and the std dev is for the annualized return i.e. geometric return. Basically for stocks the std dev of geometric return falls as time period increases. (the main idea of Jeremey Siegel's Stock for the long run, where as it increases for bonds)
Are you saying that the long term investment stochastically dominates (in the sense of first order stochastic dominance) the short-term investment? What has been highlighted in bold above suggests so.
No. Note the big assumption regarding continuity of the stock market i.e. we have assumed away a big risk of the stock market that it does continue to operate and economic growth continues. This might seem silly but many stock markets have indeed been wiped out in the last 100 years. This kind of risk does increase when you increase the time horizon. Think of it this way, is the probability of a bad economic event more likely in the next 1 hour or more likely over the next 100 years? Second assumption is regarding how we think of risk. The assumption is we think of risk in terms of portfolio falling below some x level rather than in terms of variance of the total portfolio which is a more accurate way of looking at risk but that is not how most people perceieve of risk.
Also, what measure of risk do Merton et al use?
Std dev of total return.
Vinod
Is equity investment safer over long term ?
Bobus,
You might ask if stocks are less risky in the long run in some sense than in the short term, where does this risk disapper? This has to do with mean-reversion with assoicated implications for portfolios.
Vinod
You might ask if stocks are less risky in the long run in some sense than in the short term, where does this risk disapper? This has to do with mean-reversion with assoicated implications for portfolios.
Vinod
-
Ghostwhowalks
- Posts: 40
- Joined: Sun Mar 04, 2007 11:42 pm
Is equity investment safer over long term ?
I also have different thoughts than the conventional wisdom. Let me see if I can articulate this well.
Most of the theory comes from the Benjamin Graham/Warren Buffet school of thought. The theory basocally says that one should invest in stocks for the long term (10 to 20 years- I agree with this) but that for anything shorter it is nothing but speculation ( I struggle with this). It is hard to srgue against tangible evidence of Mr Buffet's wealth but I do think there has to be some context around this. Warren became rich from investments he made in the last era- While his principles on cash flow etc are timeless, I do think that some other theries need revisiting. Here is my rationale on this:
1. Companies today are far more efficient than they were before. With real time market /operational information, they move quickly and are able to adjust to what is needed
2. Mangement takes higher risks - While it waits to churn out the next ipod, it is able to mitigate the downside by tracking the numbers closely and pulling the plug if need be, Thus it saves them from spending a lot of time and resources
3. Management churn is much more. We cannot expect the leaders today to be running the same companies 20 years from now.
4. Markets have expanded way beyond the geographies the companies are located in. This provides excellent buffer
what is the ramification of all this? In my opinion, sucess/failure cycles are much shorter now, Companies today go thru up-down-up in 3 to 5 years than 5-7-10 years. The reason why 3 to 5 years was not advocated at that time is because the cycles took longer. Now that , these have been compressed, 3 to 5 years is not "short term: anymore and therefore not as risky. 09/11 is aclassic case in point- The recovery here was in 3 to 4 years.
With this background, I feel that by looking at the history of the stock/MF, the "manager" and closely tracking it, 3 to 5 years is a good deal. I do think tat 1 to 2 years is still risky and that cycles are not this short.
Most of the theory comes from the Benjamin Graham/Warren Buffet school of thought. The theory basocally says that one should invest in stocks for the long term (10 to 20 years- I agree with this) but that for anything shorter it is nothing but speculation ( I struggle with this). It is hard to srgue against tangible evidence of Mr Buffet's wealth but I do think there has to be some context around this. Warren became rich from investments he made in the last era- While his principles on cash flow etc are timeless, I do think that some other theries need revisiting. Here is my rationale on this:
1. Companies today are far more efficient than they were before. With real time market /operational information, they move quickly and are able to adjust to what is needed
2. Mangement takes higher risks - While it waits to churn out the next ipod, it is able to mitigate the downside by tracking the numbers closely and pulling the plug if need be, Thus it saves them from spending a lot of time and resources
3. Management churn is much more. We cannot expect the leaders today to be running the same companies 20 years from now.
4. Markets have expanded way beyond the geographies the companies are located in. This provides excellent buffer
what is the ramification of all this? In my opinion, sucess/failure cycles are much shorter now, Companies today go thru up-down-up in 3 to 5 years than 5-7-10 years. The reason why 3 to 5 years was not advocated at that time is because the cycles took longer. Now that , these have been compressed, 3 to 5 years is not "short term: anymore and therefore not as risky. 09/11 is aclassic case in point- The recovery here was in 3 to 4 years.
With this background, I feel that by looking at the history of the stock/MF, the "manager" and closely tracking it, 3 to 5 years is a good deal. I do think tat 1 to 2 years is still risky and that cycles are not this short.
Is equity investment safer over long term ?
Ghostwhowalks;35285Desi
1. Can you explain your rationale of not getting into the MF market at all(active/passive is irrelevent) for 3 to 5 years?
[/quote]
Note: The discussion below is without taking inflation into account.
High risk - that is the rationale.
Let us work from some examples before we get specific to your case.
Let us say, I have saved $30,000 for a specific target to buy Toyota Camry as my old car is 12 years old and breaking down often.
Now, I have two more weeks to go before I actually buy the car. Should I put $30,000 targetted for the car into the equities market (mutual funds or stocks or options)?
No. Absolutely not.
Why not?
Because there is a finite probability that I might lose a portion of that money that I need for the car. This unnecessary risk for 2 weeks is akin to gambling and not investing and should the risk materialize of which a finite probability exists, I will not be able to achieve my goal of purchasing the car because I gambled to my detriment.
What risk of loss and how do I say this?
Largest one day losses in US markets.
Dow Jones dropped almost 23% in one day in October 1987.
Over a 2 day period in 1929 Dow Jones dropped almost 25% and one week later it dropped another 10%.
By the data I looked at (1946 to 1996), worst annualized decline for equities over five years is -2.4% annualized. This is about a 15% loss after five years.
The above data does not include the 2000 to 2005 period which I believe drags the -2.4% further and makes it worse.
I can pretty much guarantee you that the data for Nasdaq is worse than data for Dow Jones presented above.
Year 1974 - SP500 is down 26%
Year 2000 - SP 500 is down over 10%
Year 2001 - SP 500 is down over 13%
Year 2002 - SP 500 is down over 23%
Year 2000 - Nasdaq is down over 39%
Year 2001 - Nasdaq is down over 21%
Year 2002 - Nasdaq is down over 31%
Over 3 year periods, SP500 has sustained losses of 40% plus.
http://www.investorsalley.com/marketcap/11-13-06/feature.html
Clearly, then putting money in market over 2 weeks is gambling and there is a high risk of loss. Putting money in the market over a longer term is actually investing and not gambling. I will not explain this as to why putting money in market over long term (10 plus years) is investing and not gambling as you have indicated you understand the long term issues.
So, if putting money in equities for two weeks carries a high risk of loss, what about 2 months, 2 years or 20 years? Where do we draw the line? The answer lies in the data.
Data shows that if I take moving five year slices in market, over say a 50 year period or a 75 year period there are instances of five years periods where even after five years of investing there is a loss. The most recent period is 2000 to 2005. Nasdaq still after 7 years is more than 50% down from its peak.
The number of periods of loss and the amount of loss both decline as we increase the duration of these slices. So we take two year moving slices of investment in market and compare them to year slices, we see that five year slices carry a lower risk than 2 years. Based on the data, five to 7 year slices gets us to a point where the risk of loss becomes significantly small so as to start considering equity investments.
Let me take your #2, a little later today (before noon).
Is equity investment safer over long term ?
Vinod #124 and #125: Thanks again.
------------------------
I see three statements from you, that I think I understand:
(a) Std dev of annualized return of equity investment falls as time period increases.
(b) Std dev of total return of equity investment rises as time period increases.
(c) Probability of negative real return from equity investment falls as time period increases.
Did I capture your statements correctly above? I suppose these are well documented empirical regularities based on historical info, right?
-------------------------------
Now to a statement that I dont understand:
The assumption is we think of risk in terms of portfolio falling below some x level ...
Say you invest $100K, then the risk that your investment would be less than $X K (i.e. some positive real return) is reduced the longer your investment horizon ...
Suppose we think of risk in the manner described in the first statement above. If the second statement above is not about first order stochastic dominance, what is it? I understand "risk" in the second statement above to mean "probability". Does the second statement hold for negative real return or does it hold only for positive real return?
-----------------------------
You might ask if stocks are less risky in the long run in some sense than in the short term, where does this risk disapper? This has to do with mean-reversion with assoicated implications for portfolios.
Is mean reversion a necessary condition for statements (a) and (c) above? In other words, if mean reversion did not hold, can (a) and (c) conceivably hold?
Is mean reversion a sufficient condition for (a) and (c)?
Is mean reversion merely an aiding factor for (a) and (c)?
------------------------
I see three statements from you, that I think I understand:
(a) Std dev of annualized return of equity investment falls as time period increases.
(b) Std dev of total return of equity investment rises as time period increases.
(c) Probability of negative real return from equity investment falls as time period increases.
Did I capture your statements correctly above? I suppose these are well documented empirical regularities based on historical info, right?
-------------------------------
Now to a statement that I dont understand:
The assumption is we think of risk in terms of portfolio falling below some x level ...
Say you invest $100K, then the risk that your investment would be less than $X K (i.e. some positive real return) is reduced the longer your investment horizon ...
Suppose we think of risk in the manner described in the first statement above. If the second statement above is not about first order stochastic dominance, what is it? I understand "risk" in the second statement above to mean "probability". Does the second statement hold for negative real return or does it hold only for positive real return?
-----------------------------
You might ask if stocks are less risky in the long run in some sense than in the short term, where does this risk disapper? This has to do with mean-reversion with assoicated implications for portfolios.
Is mean reversion a necessary condition for statements (a) and (c) above? In other words, if mean reversion did not hold, can (a) and (c) conceivably hold?
Is mean reversion a sufficient condition for (a) and (c)?
Is mean reversion merely an aiding factor for (a) and (c)?
Is equity investment safer over long term ?
Adding more data:
Risk is defined in different ways. It is a probability of loss of your original investment ( I know that inflation can be put into this mix or kept out).
Risk can also be uncertainity of earning your investment return. Risk is classically measured as standard deviation of an investment returns. Beta as used in investments measures volatility of an investment with respect to volatility of SP500.
When talking about risk people talk about risk tolerance. Risk tolerance is not just risk psyche, but it should encompass risk afforability and risk necessity also. In other words should the risk materialize, can I afford it or would I go bankrupt and be in the poor house? and do I have any compelling need to take this risk?
Some more data:
Risk is defined in different ways. It is a probability of loss of your original investment ( I know that inflation can be put into this mix or kept out).
Risk can also be uncertainity of earning your investment return. Risk is classically measured as standard deviation of an investment returns. Beta as used in investments measures volatility of an investment with respect to volatility of SP500.
When talking about risk people talk about risk tolerance. Risk tolerance is not just risk psyche, but it should encompass risk afforability and risk necessity also. In other words should the risk materialize, can I afford it or would I go bankrupt and be in the poor house? and do I have any compelling need to take this risk?
Some more data:
On 10/09/2002, Nasdaq bottomed at 1114
On 03/10/2000, Nasdaq peaked at 5039
Loss of 78% in 31 months.
On 10/09/2002, SP500 bottomed at 776.76
On 03/24/2000, SP500 peaked at 1527.46
Loss of 49% in 30.5 months.
During September 1929, DJIA had peaked at a level of 386. By 1932, the DJIA was at a level of 40.5, that is a whopping drop of 90%.
The DJIA did not recover back to the level 386 till 25 years later in November 1954.
-
Ghostwhowalks
- Posts: 40
- Joined: Sun Mar 04, 2007 11:42 pm
Is equity investment safer over long term ?
Desi
I undersatnd this. Refer to my latest post as well since I think the market dynamics is being changed.
The Dow/Nasdaq drop crica 09/11 is mostkt because the market was irrational. They deviated far away from fundamentals of the factthat cash is still the king. People have wisened up now and I do not think we will make the same mistake again.
The problem is that post 09/11 "history is still being made" and so support from data will not be as strong. Like I say in my post, data from earlier days while still relevent to some extent needs to be put in the proper context
I undersatnd this. Refer to my latest post as well since I think the market dynamics is being changed.
The Dow/Nasdaq drop crica 09/11 is mostkt because the market was irrational. They deviated far away from fundamentals of the factthat cash is still the king. People have wisened up now and I do not think we will make the same mistake again.
The problem is that post 09/11 "history is still being made" and so support from data will not be as strong. Like I say in my post, data from earlier days while still relevent to some extent needs to be put in the proper context