200-day moving average market timing

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
BlueEars
Posts: 3968
Joined: Fri Mar 09, 2007 11:15 pm
Location: West Coast

Post by BlueEars »

Here is my list again for MT. Thanks to kencc for his inputs.

Code: Select all

Positives 
1) partial protection in bear mkts, especially slow rolloff types -- sleep well factor 
2) does not involve valuation judgements, can be purely mechanical 
3) well defined entry/exit points
4) gives one more confidence that he can take on moderate risk portfolio as he ages 
5) can mix with buy-hold to probably reduce portfolio std deviation
6) can have positive tracing error (as well as negative) for long periods 

Negatives 
1) cannot protect against very sharp drop offs (like 1987) 
2) somewhat an act of faith, no supporting underlying academic theory or "respected" proponents 
3) probably will return less then buy-hold if there is no severe market declines 
4) mechanics more difficult then buy-hold (but only ~ 1 transaction per year) 
5) possible negative tracking error with buy-hold for periods as long as 10 years 
6) possibly some MT techniques suffer from datamining
Any more suggestions for changes?
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T »

.
Data for the longest period I have (data excludes tax impacts).

Code: Select all

US SMALL VALUE - 1927/10 to 2008/11

                             Annualized Return     Growth $1

Buy and hold                      14.48              57,610
Market timing (16 month MA)       13.26              24,272

Source: Small value series from the FF 2x3 file on Ken French’s website

This is why I’m also cautious of any sweeping statements on the superiority of marketing timing over buy-and-hold. Backtest results are sensitive to asset class used, time period, duration of moving averages, buy and sell bands around the MA, and taxes (perhaps more). Of all the marketing timing back-tests I have done, the results personally don’t change my top four asset allocation priorities (for better or worse):
  • 1. Diversify across risk factors (market, size, value, term)
    2. Diversify globally (US, Non-US, EM)
    3. Minimize costs (including taxes)
    4. Rebalance (using rebalancing bands).
I will continue to track performance of some MA approaches, which may inform rebalancing frequency, but the more I look at the data, the more doubtful I become (particularly for a value and small cap tilted portfolio). Time will tell.

Just my personal view.

Robert
.
JOJO123
Posts: 337
Joined: Thu Dec 11, 2008 10:29 pm

Post by JOJO123 »

******
Last edited by JOJO123 on Mon Dec 23, 2019 10:45 pm, edited 1 time in total.
idoc2020
Posts: 1090
Joined: Mon Oct 22, 2007 3:40 pm

Post by idoc2020 »

Robert T wrote:.
Data for the longest period I have (data excludes tax impacts).

Code: Select all

US SMALL VALUE - 1927/10 to 2008/11

                             Annualized Return     Growth $1

Buy and hold                      14.48              57,610
Market timing (16 month MA)       13.26              24,272

Source: Small value series from the FF 2x3 file on Ken French’s website

This is why I’m also cautious of any sweeping statements on the superiority of marketing timing over buy-and-hold. Backtest results are sensitive to asset class used, time period, duration of moving averages, buy and sell bands around the MA, and taxes (perhaps more). Of all the marketing timing back-tests I have done, the results personally don’t change my top four asset allocation priorities (for better or worse):
  • 1. Diversify across risk factors (market, size, value, term)
    2. Diversify globally (US, Non-US, EM)
    3. Minimize costs (including taxes)
    4. Rebalance (using rebalancing bands).
I will continue to track performance of some MA approaches, which may inform rebalancing frequency, but the more I look at the data, the more doubtful I become (particularly for a value and small cap tilted portfolio). Time will tell.

Just my personal view.

Robert
.
HI Robert,
Thanks for the research. I find it quite surprising that the annual difference was only 1% but the value of $1 grew to more than double with B&H as compared to market timing. Incidentally, one of the problems of this thread is that many of us have backtested various MA strategies, with various bands, with various funds, over varying periods and it gets a little confusing. Data presented on this thread by other posters seems to confirm an approximately 1% annual underperformance over time, no matter which MA is used. I would say that if this were the price for greatly decreased volatility I would be willing to pay it. The only thing left to thesh out is which indicators we want to use for the various funds that we follow; and what bands (if any) we want to implement.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T »

FD wrote:Nice try, soon you are going to show me small index in South Africa.
  • ??
My portfolio has a target small cap and value tilt (not TSM), so was interested to see how the MA performed with SV over the 1927-08 period (to get closer to an apples to apples comparison).

For another post, I backtested monthly moving average variants from 2 to 24 months, with various buy-sell cross-over bands. 16 month with no band worked the best in the backtest from 1927-2008/10. That is what I used (only have daily data to 1966 on the Fama-French webiste which prevents backtesting over the great depression period, so used the monthly MAs). Nothing more, nothing less. Your response seems to reiterate that results are specific to asset class (TSM) and to method (200 day MA, 50/200 EMA), which was my earlier point.

Here was my earlier take (from another thread) with the daily data from Ken French's website. http://www.bogleheads.org/forum/viewtop ... ht=#340224

__________________________

ilan1h,
I find it quite surprising that the annual difference was only 1% but the value of $1 grew to more than double
That's the power of compouding. Here's a simple calculator that can demonstrate this impact (if you put in a start of $1, 81 yrs to grow, at the different returns (interest rates) above, you should get similar results (not exact because the earlier time period was not exactly 81 yrs).

http://www.moneychimp.com/calculator/co ... ulator.htm

Robert
.
Last edited by Robert T on Mon Dec 22, 2008 12:51 pm, edited 1 time in total.
User avatar
HomerJ
Posts: 21240
Joined: Fri Jun 06, 2008 12:50 pm

Re: Status of this thread

Post by HomerJ »

Dan Kohn wrote:What you need to understand is that the 200 DMA strategy (and, indeed, any market timing strategy) can kill you on both the upside and the downside. It's not hard to imagine a new bear market that completely slaughters those using this strategy. All it would take is a really drastic fall that occurred while you were in the market (such as from a horrible news event). Then, you could have a short number of drastic increases followed by slow declines, such that you were always out of the market when the big increases came.

The fact that this last bear market would have been avoided by 200 DMA tells you nothing. That's what the complaints about data mining are trying to communicate. Because there is no guarantee whatsoever that the next bear market will look anything like this. That's what a random walk means.
paddyshack wrote:What I have learnded from the 2008 market however is that I may not have the intestinal fortutude to be a B&H investor over the long haul. My conclusion is that I would be willing to sacrifice a higher return for less emotional turmoil in bear markets.
This just says that you need to hold a higher allocation of bonds. That's the way to deal with losses. Instead, you are pinning your hopes on a mechanical formula that not only may not protect you from the next big decline, but is very, very likely to cause you to miss the next big run ups.
Sorry to offer nothing new, but this is a great post.
JOJO123
Posts: 337
Joined: Thu Dec 11, 2008 10:29 pm

Post by JOJO123 »

******
Last edited by JOJO123 on Mon Dec 23, 2019 10:45 pm, edited 1 time in total.
User avatar
HomerJ
Posts: 21240
Joined: Fri Jun 06, 2008 12:50 pm

Post by HomerJ »

FD wrote:We are talking about
2. SMA of 200 days or EMA 50/200 (not 16 months)
This is my biggest problem with this "theory"... If market smoothing works by using a moving average, then it should help no matter what duration average you use...

Backtesting and Backtesting with different MAs until you find that works great with past history (i.e. 200 day MA) is just data-mining. We have no idea if that time period will work going forward.

It's an arbitrary number that fits the past data curve well. I'm sure someone in the 70s backtested with various MA strategies until they found a good number (Hey guys, look at the returns a 14-week moving average gives from 1910-1972!), and then did not do very well over the next 30 years.

Look, straight buy and hold gives you the market average. Remember, that HAS been 9%-10% which INCLUDES 50% down markets like we're experiencing today... Buying and holding through this bear market should give 9%-10% (I'll agree that past performance does not equal future results - so who knows if the 9%-10% market average will hold up for the next 30 years)

But B&H includes these horrible bear markets... They are part of the package, and still in the past B&H returned good numbers...

And that's straight B&H... Add in some diversification and rebalancing, and now you get the benefit of forcing yourself to sell high and buy low.

Hundreds of thousands of investors have tried to come with systems to "beat the market"... I'll just be happy getting the market average... You might "beat the market"... you might even have a 90% chance to "beat the market"... but you might do worse than the market too. It IS possible for a market to appear over the next 30 years where the 200 day MA does worse than the market.

I'll take the for sure market average, thank you.. No idea if that will be 5% or 10%, but I'll take the average. Too many people have tried to beat the market and lost.

I'll live below my means, save as much as I can, take the market average, and if I can even get 7% over the next 20-25 years, I get to retire and retire financially secure. Why shoot for more? Maybe I try a system and retire in 10-15 years... but maybe it will fail and I'll still be working at 80....

I'll take the market average.
User avatar
BlueEars
Posts: 3968
Joined: Fri Mar 09, 2007 11:15 pm
Location: West Coast

Post by BlueEars »

FD wrote:How the SP500 (or the DOW) did from 1927 to 2008/11 using B&H compare to EMA 50/200 (or just simple MA 200) for the same time.
I agree the 50/200 EMA looks quite attractive in backtests and beats other approaches I've briefly looked at over critical time periods. We just have to admit though that this is purely backtested results and is more subject to datamining charges then, for instance, French Fama data and analysis. In using it you will have to admit to yourself that it could fail going forward. Look at the 200 DMA analysis by Siegel I mentioned in previous post (using DJIA and +-1% bands). He had long periods of underperformance -- large tracking errors. Can you live with this? We all have to think about this given our own psychology. The Merriman articles I mentioned go into how they use a combo of MT and buy-hold. Their MT is split into 4 separate methods because they know that this stuff fails to work at times.

I'm really trying to be as completely objective as I can be on this. I'd like to be able to reduce risk with part of the portfolio beyond the diversifiers of bonds (TIPS, Ibonds in my case, maybe short term Treasuries when out of TIPS). To me other equity classes are potential equity enhancers but their correlations are too high in major down markets.
JOJO123
Posts: 337
Joined: Thu Dec 11, 2008 10:29 pm

Post by JOJO123 »

*****
Last edited by JOJO123 on Wed Dec 18, 2019 10:13 pm, edited 1 time in total.
User avatar
HomerJ
Posts: 21240
Joined: Fri Jun 06, 2008 12:50 pm

Post by HomerJ »

FD wrote: MA is not about beating B&H it's all about lowering risk. This thread established already that in extreme bull market B&H will perform better (about 40% of the time in 100 years) but in a "normal" and bear market MA is better.
Okay, fine... You manage risk buy using MA, and I'll manage risk by increasing my bond/cash allocation as I get closer to retirement.

B&H IS risky over short-time periods... and one does have to manage risk as one gets closer to retirement.
1. Stocks do better than bonds in the long run.
Do you think you need to have bonds in your portfolio?
Two reasons, to be able to rebalance, and to reduce short-term risk.
2. Are you sure B&H will produce 10% annually in the next 20 years?
Nope...
3. What do you say to a retiree that invest 50/50 and lost 20-25% in one year?
Many investors here thought that bonds will eliminate big losses just to find this year it was not true.
I'd say their asset allocation plan worked... Market dropped 40%-50%, they only lost 20%-25%. Good thing they managed risk being 50/50. If they couldn't afford to lose 25%, then they should have been 25/75.
4. Small cap do better than big cap in the long run.
Do you think you need to have big cap in your portfolio?
Yes. Just own the entire market and get small and big.
I agree with many here. This is new to me too. I based everything on back testing BUT B&H is also based on back testing.
We did check 50, 60 and 100 years for MA..how many more years we need to check?
I'd listen to you more if you weren't using managed funds that you chose for their historical performance, which almost always fails, yet I'm sure you think that part of your "system" is very smart too.

I'm not against the 200 MA. As presented here, it makes a lot of sense... But I'll still avoid it. Every system in history looked great to the ones who used it, right up until it failed. Every system in history was backtested, and looked like a winner, right up until it failed.

However, the 200 MA is pretty conservative... It probably reduces risk... and it may give you better returns OR worse returns than the market average.

I'll take the market average, which INCLUDES occasional 50% bear markets.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Post by Rodc »

Quote:
3. What do you say to a retiree that invest 50/50 and lost 20-25% in one year?
Many investors here thought that bonds will eliminate big losses just to find this year it was not true.


I'd say their asset allocation plan worked... Market dropped 40%-50%, they only lost 20%-25%. Good thing they managed risk being 50/50. If they couldn't afford to lose 25%, then they should have been 25/75.
Moreover, I'd add that if this is too big a loss to endure, rather than using what may or may not really mitigate risk like Moving average, such an investor ought to seriously consider an annuity which while not perfect (what is?) is still the best bet going if you really want to control investing risk.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Post by retired at 48 »

rrosenkoetter wrote:
This is my biggest problem with this "theory"... If market smoothing works by using a moving average, then it should help no matter what duration average you use...

Backtesting and Backtesting with different MAs until you find that works great with past history (i.e. 200 day MA) is just data-mining. We have no idea if that time period will work going forward.

It's an arbitrary number that fits the past data curve well. I'm sure someone in the 70s backtested with various MA strategies until they found a good number (Hey guys, look at the returns a 14-week moving average gives from 1910-1972!), and then did not do very well over the next 30 years.
No. The various backtesting shows the Moving Averages are robust over a wide range of time periods, duration, and buy/sell triggers. That optimization is done, showing which ones are best, is addressing a logical question of which one to use. There is no data mining...the hypothesis was set at the start.

You stated You're sure someone tested it in the 1970's, until they found a good number.

I'm sure it was tested in the 1960's , 1970's, 1980's 90's and 2000's, with good results. Even tadamsmar, a skeptic, provided links to about 50 studies, which showed moving averages work. And most of those studies dealt with individual stocks, and imposed very high 1 to 2 percent commission drags, both ways, which don't exist today.

You further state:
Hundreds of thousands of investors have tried to come with systems to "beat the market"...

Really...whats the source for these type statements?

What I have shared is I have heard these cliches for over 40 years, but one, and only one, indicator clearly beat buy and hold, namely, moving averages. No surprise to me that all the backtesting, by many, many people verified this.

Now one can have lot's of rationales and opinions that they choose to not use the tool, or ignore it, or it is not worth it...which is fine. But the day has arrived that people need to stop dancing around the data and face the music.

It is why one outside agency (skeptics) did a study, confirming moving averages work, stating:
The most important point of our analysis is: A simple, mechanical moving average system beats the Buy & Hold of an index fund over 20 year periods and with 30% less risk.
And asked:

...Why do so many commentators and so-called experts continually bash market timing(moving averages) when it obviously works?

There comes a time when the data and backtesting stands on its own. I submit we are there.

What one does with this is a matter of broad discussion and varied viewpoints.

R48
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Post by tadamsmar »

MA is not about beating B&H it's all about lowering risk. This thread established already that in extreme bull market B&H will perform better (about 40% of the time in 100 years) but in a "normal" and bear market MA is better.
Here we go again.

If MA is about beating the risk-adjusted return of a portfolio on the efficient frontier, then its about beating B&H.

If you accept that you can't beat an efficient portfolio, then you can just adjust risk by moving toward the less risky end of the frontier. That is, increase your bond allocation. No need to do anything more complex.

We have been around this block on this thread before.
User avatar
BlueEars
Posts: 3968
Joined: Fri Mar 09, 2007 11:15 pm
Location: West Coast

Post by BlueEars »

FD wrote:...I'm getting tired.

You guys just disregard some info.
We're just a critical bunch. Please don't take the critiques of MA personally. Also after 10 pages of this, most people will forget some of the info in previous postings.
MA is not about beating B&H it's all about lowering risk. This thread established already that in extreme bull market B&H will perform better (about 40% of the time in 100 years) but in a "normal" and bear market MA is better.

IMO, EMA 50/200 is best for me since I don't want to buy/sell more than once per year on average.
I agree with this. If I do choose to go with some MA approach for part of the portfolio then EMA 50/200 looks good.
Please answer a few questions..

1. Stocks do better than bonds in the long run.
Do you think you need to have bonds in your portfolio?
Yes, particularly bonds that will do OK in an inflationary period.
2. Are you sure B&H will produce 10% annually in the next 20 years?
Probably good odds on this now.
3. What do you say to a retiree that invest 50/50 and lost 20-25% in one year?
It's too late for regrets. You can only invest going forward. See my answer to #2 above.
4. Small cap do better than big cap in the long run. Do you think you need to have big cap in your portfolio?
I cannot handle big negative tracking errors -- that's me for better or worse. So I do not value tilt. Have plenty of big caps in US + international.
User avatar
HomerJ
Posts: 21240
Joined: Fri Jun 06, 2008 12:50 pm

Post by HomerJ »

retired at 48 wrote:
rrosenkoetter wrote: Hundreds of thousands of investors have tried to come with systems to "beat the market"...

Really...whats the source for these type statements?
Hmm.. I don't know... how many people have worked in the mutual fund industry over the last 40 years? How many people have tried to invest for themselves or for others over the last 200 years?

How many of them have consistently "beat the market"?

If MA works all the time, and has been known about for decades, how come we haven't seen some mutual funds beating the market year after year for the last 30 years?

Heck, if it actually worked, and mutual funds and money managers starting using it, soon EVERYONE would be using it.

Or are you guys saying that ONLY you and few select people on other boards know about this incredible fool-proof backtested investing system?

(Sorry to be rude. I actually enjoy most of your posts R48, and I understand your backtesting results... But you KNOW people in the past have used systems that backtested perfectly, and still failed going forward)
User avatar
BlueEars
Posts: 3968
Joined: Fri Mar 09, 2007 11:15 pm
Location: West Coast

Post by BlueEars »

rrosenkoetter wrote:...(Sorry to be rude. I actually enjoy most of your posts R48, and I understand your backtesting results... But you KNOW people in the past have used systems that backtested perfectly, and still failed going forward)
Perhaps you are not so sorry to be rude? Wouldn't it be great if we could keep these threads more polite and show some respect for others we do not agree with. Your points would count just as much and probably more.

Part of the reason money managers may not use moving averages too directly is because of long periods of possible negative tracking error. They would not have a job after a few years. Also fund managers would get little respect if they advertised their use of a pure moving average approach. See this article for some discussion: Which is better, buy-and-hold or market timing? link: http://www.fundadvice.com/fehtml/bhstrategies/9712.html
JOJO123
Posts: 337
Joined: Thu Dec 11, 2008 10:29 pm

Post by JOJO123 »

******
Last edited by JOJO123 on Mon Dec 23, 2019 10:44 pm, edited 1 time in total.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

200 MA

Post by retired at 48 »

rrosenkoetter wrote:
retired at 48 wrote:
rrosenkoetter wrote: Hundreds of thousands of investors have tried to come with systems to "beat the market"...

Really...whats the source for these type statements?
Hmm.. I don't know... how many people have worked in the mutual fund industry over the last 40 years? How many people have tried to invest for themselves or for others over the last 200 years?

How many of them have consistently "beat the market"?

If MA works all the time, and has been known about for decades, how come we haven't seen some mutual funds beating the market year after year for the last 30 years?

Heck, if it actually worked, and mutual funds and money managers starting using it, soon EVERYONE would be using it.

Or are you guys saying that ONLY you and few select people on other boards know about this incredible fool-proof backtested investing system?

(Sorry to be rude. I actually enjoy most of your posts R48, and I understand your backtesting results... But you KNOW people in the past have used systems that backtested perfectly, and still failed going forward)
I guess I can only keep repeating, and repeating, and repeating:

MUTUAL FUND MANAGERS HAVE TO BUY STOCKS...BOGLEHEADS BUY MUTUAL FUNDS...HUGE DIFFERENCE!

There are a variety of reasons fund managers can't succeed in beating each other, but the most obvious is that THEY ARE THE MARKET. THEY EARN THE MARKET RETURN, LESS COSTS. THIS IS A MATHEMATICAL TRUISM.

You, as an individual investor, have many unfair advantages compared to mutual fund managers. They can't take money off the table at highs (as huge money flows in); you can. They can't simply exit, selling huge numbers of stock shares, just when a 200 day MA is triggered; you can quickly sell funds. They can't go to another portfolio with the click of a mouse; you can...and on and on.

But when one says there are some periods when MA is behind Buy/Hold, these are all strong up markets. So let's say we had a 50% up market in one year. And Buy and Holders returned 49.65%...excellent. That buy and holders (on average) in active funds returned 48%...yes, a little less. That MA may have only returned 40%...so what? One doesn't "lose", in whipsaws, you lose opportunity for maximum gain.

What one clearly gets, though, is avoidance of huge downside losses. And when I combine this with the Pyramid Up Buying (By definition, continued buying after gains only), losses are even further mitigated...by definition.

This is what I tried to do to meet W. Buffett's Rule #1: Don't lose money! And I think his rule #2 was to reread Rule #1.

Just things to ponder.

BTW rrosenkoetter...a sincere thanks for the comment you enjoy most of my posts :!:

retired at 48
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: 200 MA

Post by Dan Kohn »

retired at 48, I believe that you are mistaken about most of your assertions in this post.
retired at 48 wrote:There are a variety of reasons fund managers can't succeed in beating each other, but the most obvious is that THEY ARE THE MARKET. THEY EARN THE MARKET RETURN, LESS COSTS. THIS IS A MATHEMATICAL TRUISM.
No, you are confusing an argument for indexing with an argument against running a successful mutual fund. Although all mutual funds on average must produce the same returns as indexed funds, there is nothing stopping any individual fund from beating the market (and on average, excluding fees, half do every year). If your 200 DMA market timing strategy worked reliably, mutual funds could and would be started around it without violating any mathematical trusim. It is only this rule in combination with the lack of predictably of future performance that mean that indexing is a better choice.
retired at 48 wrote:You, as an individual investor, have many unfair advantages compared to mutual fund managers. They can't take money off the table at highs (as huge money flows in); you can. They can't simply exit, selling huge numbers of stock shares, just when a 200 day MA is triggered; you can quickly sell funds. They can't go to another portfolio with the click of a mouse; you can...and on and on.
This is all completely incorrect. Mutual funds can and do move to cash whenever their manager's feel like it. Many times a month in some cases.

Huge new inflows in no way prevents a mutual fund manager from staying in cash if that is what their system is telling them to do.

Also, as long as you are talking about the S&P 500, the market for each of these stocks is so liquid that people can and do move many billions of dollars in and out every day without moving the price more than a couple pennies either way. If you wanted to start your 200 DMA fund, it would take many decades of amazing returns before you were even close to having an impact on the market.
retired at 48 wrote:MUTUAL FUND MANAGERS HAVE TO BUY STOCKS...BOGLEHEADS BUY MUTUAL FUNDS...HUGE DIFFERENCE!
Even this is wrong, since mutual funds can and do trade ETFs like SPY rather than holding individual shares. The market for SPY is one of the most liquid in the world. $21 B were traded today alone.
retired at 48 wrote:But when one says there are some periods when MA is behind Buy/Hold, these are all strong up markets. So let's say we had a 50% up market in one year. And Buy and Holders returned 49.65%...excellent. That buy and holders (on average) in active funds returned 48%...yes, a little less. That MA may have only returned 40%...so what? One doesn't "lose", in whipsaws, you lose opportunity for maximum gain.

What one clearly gets, though, is avoidance of huge downside losses.
Your examples are deeply misleading. Any market timing rule whatsoever will have reduced risk from buy and hold, by definition, since you will have time out of the market when you are not taking any risk. But in exchange, you are giving up significant returns. It's easy to construct a market scenario in which your strategy will cause no mitigation of losses whatsoever. A big enough terrorist attack could cause a 50% fall in prices before you have the time to exit the market. There is also no guarantee whatsoever that your strategy will participate in the upside, since rapid increases followed by slow downward movement will mean that you are always out of the market when the increases take place.
retired at 48 wrote:This is what I tried to do to meet W. Buffett's Rule #1: Don't lose money! And I think his rule #2 was to reread Rule #1.
This quote shows a deep misreading of Buffett's philosophy. I was at the Berkshire Hathaway annual meeting in Omaha this year where, in answer to an audience question, he recommended Vanguard funds instead of trying to beat the market. He does not complain when the prices of his investments fall, and he certainly does not bail out from his holdings (since his preferred holding time, like Jack Bogle's, is forever). He certainly does not believe that some mechanistic approach to market timing, or any technical analysis whatsoever for that matter, is a reasonable way to approach the market.
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Reading request

Post by Dan Kohn »

retired at 48, could I request that before we spend too much more time discussing your market timing approach, that you read (or re-read) Chapter 7 of Professor Burton Malkiel's magisterial (yet very readable) book, A Random Walk Down Wall Street. Malkiel is the person who convinced Jack Bogle to start the first mainstream index fund, and he still serves on Vanguard's board. Conveniently, the chapter is available for free from Google Books: http://books.google.com/books?id=0uEcVn ... #PPA144,M2

This may give you some idea of the intellectual tide you are swimming against in arguing that a market timing strategy will be successful. Yes, of course, you have found one that is successful as measured with today's market nadir as the endpoint. But the chapter can hopefully help you see how unlikely it is that a 200 DMA (or any other market timing approach) will continue to beat buy and hold on a risk-adjusted basis over time.

Here's a quote from the chapter:
Professor H. Negat Seybun of the University of Michigan found that 95 percent of the significant market gains over the thirty-year period from the mid-1960s through the mid-1990s came on 90 of the roughly 7500 trading days. If you happened to miss those 90 days, the generous long-run stock market returns of the period would have been wiped out.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Reading request

Post by Rodc »

Dan Kohn wrote:retired at 48, could I request that before we spend too much more time discussing your market timing approach, that you read (or re-read) Chapter 7 of Professor Burton Malkiel's magisterial (yet very readable) book, A Random Walk Down Wall Street. Malkiel is the person who convinced Jack Bogle to start the first mainstream index fund, and he still serves on Vanguard's board. Conveniently, the chapter is available for free from Google Books: http://books.google.com/books?id=0uEcVn ... #PPA144,M2

This may give you some idea of the intellectual tide you are swimming against in arguing that a market timing strategy will be successful. Yes, of course, you have found one that is successful as measured with today's market nadir as the endpoint. But the chapter can hopefully help you see how unlikely it is that a 200 DMA (or any other market timing approach) will continue to beat buy and hold on a risk-adjusted basis over time.

Here's a quote from the chapter:
Professor H. Negat Seybun of the University of Michigan found that 95 percent of the significant market gains over the thirty-year period from the mid-1960s through the mid-1990s came on 90 of the roughly 7500 trading days. If you happened to miss those 90 days, the generous long-run stock market returns of the period would have been wiped out.
Unfortunately we've picked that particular piece of evidence apart (and it was mainstream bogleheads who picked it apart first). The problem is extreme days cluster. It is highly unlikely any strategy, market timing or otherwise, will ever manage to catch or miss a large number of great or horrible days without also catching or missing the horrible or great days. It is not so much that that particular statistic is wrong and it is simply not relevant.

I'm on your side, but unfortunately that particular argument does not hold up.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
BlueEars
Posts: 3968
Joined: Fri Mar 09, 2007 11:15 pm
Location: West Coast

Re: Reading request

Post by BlueEars »

Dan Kohn wrote:retired at 48, could I request that before we spend too much more time discussing your market timing approach, that you read (or re-read) Chapter 7 of Professor Burton Malkiel's magisterial (yet very readable) book, A Random Walk Down Wall Street. ...
I'm not R48 but will make some observations anyway. I am trying to be open minded about some of the MT ideas addressed in this thread. Malkiel does not seem to really provide much evidence against the moving average approachs mentioned on this thread. He says:
Do you think that a moving average system ... can lead to extraordinary profits? Not if you have to pay transaction charges -- to buy and sell.
First, no one (I think) is discussing this in the context of extraordinary profits. Rather as a risk reduction technique to capture a high percentage of buy-hold returns. Second, transaction costs are miniscule on very liquid ETF's and there are none on no-load funds. When Malkiel first published this book there were probably higher transaction costs.

Here are some numbers from Siegel's study (see my reference in previous posts) of 200 DMA on DJIA with +-1% bands (numbers are return/stddev):

Code: Select all

1926-1945  buy-hold  6.3/31.0  MT  11.1/21.8
1946-2001  buy-hold 11.5/16.2  MT  10.7/14.2
1990-2001  buy hold 14.1/14.3  MT   7.4/18.4
Note the 1990-2001 showed high negative tracking error and less risk reduction. Of course, the 2003-2008 data that is missing would have made up for all this. But would an investor have been able to tolerate the tracking error? If employing an MT technique you have to answer this for yourself up front and remember it for perhaps a decade. I keep harping on this because it's easy to dismiss after a clear victory for MT this year.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Re: Reading request

Post by retired at 48 »

Dan Kohn writes:
retired at 48, could I request that before we spend too much more time discussing your market timing approach, that you read (or re-read) Chapter 7 of Professor Burton Malkiel's magisterial (yet very readable) book, A Random Walk Down Wall Street. Malkiel is the person who convinced Jack Bogle to start the first mainstream index fund, and he still serves on Vanguard's board. Conveniently, the chapter is available for free from Google Books: http://books.google.com/books?id=0uEcVn ... #PPA144,M2
Wrong study. I have indeed read many, many books on these subjects, including this one. And No, Chapter 7 is irrelevant because it deals with buying individual stocks. And whose against index funds? I suspect you haven't read all the posts and threads on these matters, for I am a strong supporter of index funds. But I also grew up with active funds, because that is all that was available, at times. For example, Emerging Markets. Because of this, that's where one learned to take insurance against complete fund blowups. A major part of my portfolio is in index funds.

This may give you some idea of the intellectual tide you are swimming against in arguing that a market timing strategy will be successful. Yes, of course, you have found one that is successful as measured with today's market nadir as the endpoint. But the chapter can hopefully help you see how unlikely it is that a 200 DMA (or any other market timing approach) will continue to beat buy and hold on a risk-adjusted basis over time.
No. The Malkiel study deals with buying individual stocks. And no, the backtesting (BTW all done by outside sources and several other Bogleheads) was over several time periods, of several lengths of time, ending at differing years. No cherry picking the current market. And you insist on calling it a timing strategy instead of an investment approach that is clearly more conservative than others, including buy and hold. Want an example. One backtester even took the Japanese Market from its peak, backtested to today (I assume you are familiar with this huge two decade decline), showing that MA did not lose. You keep assuming someone who just inherited $500,000 cash wants to maximize their forward going returns. I submit their real goal is "good heavens, I don't want to lose 25% of this, or my dad will roll over in his grave." Academic studies can't handle this.

Here's a quote from the chapter:

Professor H. Negat Seybun of the University of Michigan found that 95 percent of the significant market gains over the thirty-year period from the mid-1960s through the mid-1990s came on 90 of the roughly 7500 trading days. If you happened to miss those 90 days, the generous long-run stock market returns of the period would have been wiped out.
Wrong again. This has been shown in other studies to be a mathematical trickery. I don't have the disputing argument at hand. But the same thing said: miss the XX number of bad days, and you earned far more. (Edit: I see where Rodc addresses this above)

BTW this year we have added new records for some of those "up days", and where are we...a market low!

But, the key is the backtesting. Surely the backteasting would have showed some huge under-performance in outcomes, if some time periods were "missing these key days." But it doesn't. Nada. You can't have this both ways. Either the backtesting shows positive results, or not.

Lastly, the approach involves individuals buying mutual funds, in a conservative manner. And it has been stated that one dilemma is academics cannot study this, because such individual results are not in ready forms of data to be studied.

R48
Last edited by retired at 48 on Mon Dec 22, 2008 5:41 pm, edited 1 time in total.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Re: 200 MA

Post by retired at 48 »

Dan Kohn" wrote: retired at 48, I believe that you are mistaken about most of your assertions in this post.
retired at 48 wrote:There are a variety of reasons fund managers can't succeed in beating each other, but the most obvious is that THEY ARE THE MARKET. THEY EARN THE MARKET RETURN, LESS COSTS. THIS IS A MATHEMATICAL TRUISM.
No, you are confusing an argument for indexing with an argument against running a successful mutual fund. Although all mutual funds on average must produce the same returns as indexed funds, there is nothing stopping any individual fund from beating the market (and on average, excluding fees, half do every year). If your 200 DMA market timing strategy worked reliably, mutual funds could and would be started around it without violating any mathematical trusim. It is only this rule in combination with the lack of predictably of future performance that mean that indexing is a better choice.

R48 reply: No. no one is suggesting starting up a 200 day MA Mutual Fund. Nor is anyone against index funds. The majority of my funds are index funds. Regarding whether professional institutional investors use these techniques, some seem to have a keen interest, as noted by this:

I received a PM from a small Asset Management/Advisory Firm located near Valley Forge, PA. Their specialty is the large cap USA stock market space. They invited me to a Q & A session regarding some of my postings. They would request all staff members read the postings “First Time Posters IRA @ $1,000,000: Shares Lessons Learned”; postings on “The Pyramid Up Buying Technique”, and selected posts regarding momentum enhanced investing.

During my transition from Florida to Saratoga, NY I visited, and received cordial treatment in a conference room by the owner and all staff members. I made a presentation; including sharing contrary Boglehead philosophy regarding active management and the potential drag the fees/costs can have on performance. They understood these positions. A good question and answer period followed.

retired at 48
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Re: 200 MA

Post by retired at 48 »

"Dan Kohn"wrote:
"retired at 48"]You, as an individual investor, have many unfair advantages compared to mutual fund managers. They can't take money off the table at highs (as huge money flows in); you can. They can't simply exit, selling huge numbers of stock shares, just when a 200 day MA is triggered; you can quickly sell funds. They can't go to another portfolio with the click of a mouse; you can...and on and on.
This is all completely incorrect. Mutual funds can and do move to cash whenever their manager's feel like it. Many times a month in some cases. Huge new inflows in no way prevents a mutual fund manager from staying in cash if that is what their system is telling them to do.

R48 reply: No. Most mutual fund managers can't go to cash. Tell that to Jeff Vinick of Fidelity's huge Magellen Fund (after Peter Lynch), who went to cash, then was fired. No matter he may have been a little early, investors fled.

Tell that to a guy name Graham, a value investor who felt the market was too high, going to cash too early, and Vanguard fired him.

And tell that to Bill Miller, who two years ago would likely have desired to trade his portfolio for another. He has spent two years selling stocks bigtime to meet heavy redemptions of those fleeing his fund.

No, most funds do not go to cash, and they can't vacate their portfolios. Sometimes they are forced into cash.

R48
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

200 ma

Post by retired at 48 »

Dan Kohn wrote:
A big enough terrorist attack could cause a 50% fall in prices before you have the time to exit the market. There is also no guarantee whatsoever that your strategy will participate in the upside, since rapid increases followed by slow downward movement will mean that you are always out of the market when the increases take place.
For a 50% fall, the default position is the same as "Buy and Hold". Each experiemces the decline. Of course, an instantaneous large down-day will affect both, except for the 20% of the time the MA investor is not in the market...an advantage.

Your description of rapid increases, followed by slow downward movements as suggesting one is always out of the market is not correct. The 200 day MA eventually catches up, and any rapid increase will cross the 200 day MA. Secondly, if one uses the Pyramid Up buying strategy, this potential dilemma is solved.

R48
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Post by retired at 48 »

Dan Kohn wrote:
retired at 48 wrote:

This is what I tried to do to meet W. Buffett's Rule #1: Don't lose money! And I think his rule #2 was to reread Rule #1.
This quote shows a deep misreading of Buffett's philosophy.

R48 Reply:

How can a quote be a deep misreading. That is the Rule he cites.

What Buffett does (like buying complete companies, and holding forever), or what you may do, or what any other investor may do, is not relevant. My comment is WHAT I DID. And many Bogleheads have stated they like to hear what others actually did, to supplement these theories.

So when I state I did such and such to give the rule meaning (such as how to not lose money), then one cannot question it, only ponder what one person did. It's not debatable, (unless one wants to challenge it as something I didn't do, i.e., untruthful).

This is how we learn...theory and actual practice.

retired at 48
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: 200 MA

Post by Dan Kohn »

retired at 48 wrote:No. Most mutual fund managers can't go to cash. Tell that to Jeff Vinick of Fidelity's huge Magellen Fund (after Peter Lynch), who went to cash, then was fired. No matter he may have been a little early, investors fled.

Tell that to a guy name Graham, a value investor who felt the market was too high, going to cash too early, and Vanguard fired him.

And tell that to Bill Miller, who two years ago would likely have desired to trade his portfolio for another. He has spent two years selling stocks bigtime to meet heavy redemptions of those fleeing his fund.

No, most funds do not go to cash, and they can't vacate their portfolios. Sometimes they are forced into cash.
By looking at 2 of the largest funds in history, you've picked very misleading examples. Further, these funds were (I believe) supposed to remain mostly invested.

I think you may be confused about the diversity of mutual funds. They buy and sell stocks in a way that is more convenient for individuals than doing it themselves. That's it. There's no more rules beyond that (well, they can't go short, and there are some reporting requirements, etc.). And if any group thought that a simple market timing of index funds strategy would beat the risk-adjusted return of a regular index fund, such a fund would certainly be created. In fact, I suspect one already exists, or did at one time.

Look at all of your followers who are looking for simple rules they can follow to reduce their losses. They would all be potential customers. Except that, as with every market timing strategy in the history of the world, yours is nearly certain to underperform the market (on a risk-adjusted basis) in the time period after you create it, and you will find less and less adherents over time.

My point stands, though, that you were incompletely incorrect in saying that you are suggesting some sort of strategy that could be followed by individual investors, but not by a mutual fund.
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: Reading request

Post by Dan Kohn »

Les wrote:
Malkiel wrote:Do you think that a moving average system ... can lead to extraordinary profits? Not if you have to pay transaction charges -- to buy and sell.
First, no one (I think) is discussing this in the context of extraordinary profits. Rather as a risk reduction technique to capture a high percentage of buy-hold returns. Second, transaction costs are miniscule on very liquid ETF's and there are none on no-load funds. When Malkiel first published this book there were probably higher transaction costs.
R48 is not necessarily promising higher returns. He's promising lower volatility. And a market timing strategy is guaranteed to provide that, since any time you are out of the market you will have zero volatility.

So the real question isn't whether we can get lower volatility, because we can all just hold T-Bills and bring volatility to zero. The question is whether R48's market timing strategy will provide higher risk-adjusted returns in time periods outside of the ones that have been back-tested for (i.e., in the future). And, it should compare not to the returns of buy and hold of an index, but of buy and hold of an asset allocation of that index and bonds. Since if you're willing to accept lower returns in exchange for lower volatility, the null hypothesis should be the use of an asset allocation with bonds to do that.

Let me lay out a backtest that I personally would find more convincing. Others on the board are in a better position to state whether this would be more statistically relevant, or whether I'm missing something. Take all 62 20 year periods from 1926 through 2008 (e.g., 1926..1945, 1927..1946, etc.). In each, calculate the return and standard deviation of DMA200 strategy, and a 100/0, 75/25, and 50/50 mix of stocks and bonds. Then, compare the number of periods where the Sharpe Ratio of the DMA 200 beats the different asset allocations. If there were something magic about DMA 200, I would expect it to have a higher Sharpe Ratio far more than 50% of the time against each of the asset allocations.

Sorry that I'm not in a position to do this analysis myself.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Re: 200 MA

Post by retired at 48 »

Dan Kohn wrote:
retired at 48 wrote:No. Most mutual fund managers can't go to cash. Tell that to Jeff Vinick of Fidelity's huge Magellen Fund (after Peter Lynch), who went to cash, then was fired. No matter he may have been a little early, investors fled.

Tell that to a guy name Graham, a value investor who felt the market was too high, going to cash too early, and Vanguard fired him.

And tell that to Bill Miller, who two years ago would likely have desired to trade his portfolio for another. He has spent two years selling stocks bigtime to meet heavy redemptions of those fleeing his fund.

No, most funds do not go to cash, and they can't vacate their portfolios. Sometimes they are forced into cash.
By looking at 2 of the largest funds in history, you've picked very misleading examples. Further, these funds were (I believe) supposed to remain mostly invested.

I think you may be confused about the diversity of mutual funds. They buy and sell stocks in a way that is more convenient for individuals than doing it themselves. That's it. There's no more rules beyond that (well, they can't go short, and there are some reporting requirements, etc.). And if any group thought that a simple market timing of index funds strategy would beat the risk-adjusted return of a regular index fund, such a fund would certainly be created. In fact, I suspect one already exists, or did at one time.

Look at all of your followers who are looking for simple rules they can follow to reduce their losses. They would all be potential customers. Except that, as with every market timing strategy in the history of the world, yours is nearly certain to underperform the market (on a risk-adjusted basis) in the time period after you create it, and you will find less and less adherents over time.

My point stands, though, that you were incompletely incorrect in saying that you are suggesting some sort of strategy that could be followed by individual investors, but not by a mutual fund.
Sorry to disagree. If you mean some mutual fund can become a fund-of-funds, investing around a 200 day Moving Average (MA) scheme, perhaps it could be tried. Perhaps it has.

But an individuals portfolio, designed for his specific needs, is what is discussed throughout this thread, This cannot be lumped into mutual funds. For example, we discussed making 401.k payroll adjustments based on MA's; rebalancing around MA's instead of birthdays; changing reallocation percentages when one ages, for example, and using 200 day MA's to make those changes. Deciding one doesn't want REIT's in their allocation, and how to exit. You receive an inheritance, and have to decide how and when to invest it.

So, no, this can be done by investment advisers like a Rick Ferri, dealing with individuals, but not en mass in a common mutual fund.

There's two parts to the thread...do MA's work, and what can one practically do to utilize this tool?

R48
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: Reading request

Post by Dan Kohn »

retired at 48 wrote:One backtester even took the Japanese Market from its peak, backtested to today (I assume you are familiar with this huge two decade decline), showing that MA did not lose.
Of course, it didn't lose. The Japanese stock market is in the worst slump in 20+ years. Any market timing strategy that pushes you out of the market is going to look superior to buy and hold, when your end point is the current nadir. And it will be superior. As long as you are ready to sell everything now and never again invest. But if you actually need to earn a reasonable return going forward, your market timing strategy is very likely going to cause your followers to miss out on the inevitable rises ahead. And they need those climbs to earn a reasonable return for their retirements.
retired at 48 wrote:You keep assuming someone who just inherited $500,000 cash wants to maximize their forward going returns. I submit their real goal is "good heavens, I don't want to lose 25% of this, or my dad will roll over in his grave."
That is a silly strawman. I'm suggesting that when investors cannot handle volatility, they should hold more bonds. Nearly everyone who is considering following your strategy will be much better off if they increase they bond holdings and stick with buy and hold for the equity portion of their asset allocation. In fact, I am suggesting that a buy and hold strategy (as part of a reasonable asset allocation) will provide lower volatility than your market timing strategy, for any given level of desired return. Phrased differently, I'm suggesting that buy and hold (with bonds) will provide the highest possible return for any acceptable level of risk. That is the message of the Random Walk Hypothesis, Efficient Market Theory, and Modern Portfolio Theory. The creators of those theories won several Nobel prizes. (We'll momentarily ignore the billions of dollars they went on to lose when they started trying to beat the market.)

Now, the Wikipedia links I provided include criticism sections describing the attempts to shoot weaknesses in these theories. But, at least there are intellectually serious efforts to explain the world.

By contrast, your market timing suggestions have no theory behind them whatsoever. They are just back testing, and doing it in a market nadir when market timing strategies are guaranteed to do well.
retired at 48 wrote:Academic studies can't handle this.... And it has been stated that one dilemma is academics cannot study this, because such individual results are not in ready forms of data to be studied.
This is a deeply silly statement. Would you like to elaborate what you meant? What is stopping academics (or us) from evaluating your market timing strategy?
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: Reading request

Post by Dan Kohn »

Rodc wrote:
Malkiel wrote:Professor H. Negat Seybun of the University of Michigan found that 95 percent of the significant market gains over the thirty-year period from the mid-1960s through the mid-1990s came on 90 of the roughly 7500 trading days. If you happened to miss those 90 days, the generous long-run stock market returns of the period would have been wiped out.
Unfortunately we've picked that particular piece of evidence apart (and it was mainstream bogleheads who picked it apart first). The problem is extreme days cluster. It is highly unlikely any strategy, market timing or otherwise, will ever manage to catch or miss a large number of great or horrible days without also catching or missing the horrible or great days. It is not so much that that particular statistic is wrong and it is simply not relevant.

I'm on your side, but unfortunately that particular argument does not hold up.
I believe the quote is highly relevant. The goal of any market timing strategy is to be in the market on the good days and out on the bad days. I believe there is very strong evidence that the days are random. But the fact that missing a small number of good days has an enormously detrimental effect on your portfolio explains why it is so hard to come up with a market timing strategy that reliably beats buy and hold across different time periods.

What the followers of this thread are looking for is a strategy that will miss more bad days than good. But such a strategy doesn't exist. Instead, they should accept the bad days with the good, and deal with the bad ones by holding more bonds.
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Post by tadamsmar »

retired at 48 wrote:
Academic studies can't handle this.... And it has been stated that one dilemma is academics cannot study this, because such individual results are not in ready forms of data to be studied.

This is a deeply silly statement. Would you like to elaborate what you meant? What is stopping academics (or us) from evaluating your market timing strategy?
That's impressive. Academics study things like prayer. MA must be more esoteric than prayer in 48's mind.

It's news to me that this is individual. I thought these system involved just follow signals calculated from trailing returns.

Much more existential and esoteric than I imagined!

So this is not a religion, its too esoteric to be a religion!
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: 200 ma

Post by Dan Kohn »

retired at 48 wrote:Dan Kohn wrote:
A big enough terrorist attack could cause a 50% fall in prices before you have the time to exit the market. There is also no guarantee whatsoever that your strategy will participate in the upside, since rapid increases followed by slow downward movement will mean that you are always out of the market when the increases take place.
For a 50% fall, the default position is the same as "Buy and Hold". Each experiemces the decline. Of course, an instantaneous large down-day will affect both, except for the 20% of the time the MA investor is not in the market...an advantage.

Your description of rapid increases, followed by slow downward movements as suggesting one is always out of the market is not correct. The 200 day MA eventually catches up, and any rapid increase will cross the 200 day MA.
R48, my point is that your market timing strategy will do well in certain kinds of declines and certain kinds of increases (generally, slower ones for both). But that in many markets (most, based on my understand of modern portfolio theory), it will do worse than buy and hold. In particular, rapid increases followed by slow declines would cause you to miss all increases.
retired at 48 wrote:Secondly, if one uses the Pyramid Up buying strategy, this potential dilemma is solved.
Could we please discuss one market timing strategy at a time? Otherwise, you could be accused of changing your recommendation post to post.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

Post by retired at 48 »

Dan Kohn wrote:

retired at 48 wrote:

Academic studies can't handle this.... And it has been stated that one dilemma is academics cannot study this, because such individual results are not in ready forms of data to be studied.
This is a deeply silly statement. Would you like to elaborate what you meant? What is stopping academics (or us) from evaluating your market timing strategy?

R48 Reply

Because studying one individuals performance is easily ascribed to "luck"....or a favorable investment time period, and on and on.

Most individuals near retirement did not save records "to be backtested." It is hearsay.

Academics would insist on at least 300 random cases, an impossibility to satisfy.

We've heard a thousand times that nobody can beat the market, but the early retirees in my community have no choice but to accept that: "it was all luck."

Lastly, I suppose you could start with this individual. His name is James Simons. He is the highest earning salary person in the world...over $1 Billion last year.

He was a professor, who felt he could develop an algorithm that could beat the market. He did. It did. He runs the most successful hedge fund on Wall Street, and has not shown Madoff tendencies. He has 200 people, daily, trading stocks. He has survived close scrutiny, returning about 24% a year for the past 15 years. Google him.

They told him it couldn't work, too, because nobody beats the market.
R48
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

My views

Post by Dan Kohn »

retired at 48 wrote:So when I state I did such and such to give the rule meaning (such as how to not lose money), then one cannot question it, only ponder what one person did. It's not debatable, (unless one wants to challenge it as something I didn't do, i.e., untruthful).

This is how we learn...theory and actual practice.
R48, I can't question your market timing theory because you don't have a theory. You have a backtest showing that DMA200 has outperformed buy and hold (with no bonds) using today as the end date. You can't explain why it did so, or under what circumstances it would continue to, or what might cause it to stop.

But let me lay my cards on the table. We shared a beer at the Bogleheads meeting in San Diego in September. I do not think that you have been untruthful. I also do not think that you are intentionally taking advantage of anyone, or that in any of your 1600 posts you have had any intention other than helping people. Nonetheless, I think you are in the process of doing an enormous amount of damage, and I have decided to spend the several dozen hours that it will probably take to try to explain why.

You were able to retire at 48, which instantly raises your credibility with other forum readers. However, you did this with help from two very non-Boglehead approaches: you leveraged up and you market timed. Now, in your case, that went great. You happened to be investing in the 80's and 90's during the greatest bull market the world has ever seen. And so your market timing signals did not leave you out of the market for very long. And your leverage increased your returns without causing your portfolio to blow up, as occurred with market timer.

But unfortunately, I believe that your successes have led you to exactly the wrong lessons about how others can best succeed. Because success should not be period-dependent. And the two strategies that you followed, if you had done them in different periods, would have led to utter disaster. That's what I don't think you understand when people are skeptical about the value of back testing.

You promote index funds, and that is an important part of Boglehead philosophy. But it may even be the least important part. Large, actively managed fund (like American funds) become closet index funds anyway, and if they were available at low cost, would be almost as good. But far more important is the idea that the way to deal with risk is through the use of asset allocation (i.e., holding bonds), not trying to time ups and downs.
Topic Author
JW-Retired
Posts: 7189
Joined: Sun Dec 16, 2007 11:25 am

Post by JW-Retired »

Robert T wrote: US SMALL VALUE - 1927/10 to 2008/11
Annualized Return Growth $1
Buy and hold 14.48 57,610
Market timing (16 month MA) 13.26 24,272
Source: Small value series from the FF 2x3 file on Ken French’s website
I'm doubting the meaning of this because you are using monthly data. This just cannot be the same as simulating using daily data. Think about it. Your monthly data based simulation is creating some signal that is in effect acted on with weeks of lag compared to when the investors on the ground at the time see the same conditions. By the time the simulation reacts it is going to be old news to the market.

If we think this MT stuff works because of some average investor psychology, then you have to simulate using the same fidelity of data that real investors are seeing.
JW
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Academics.

Post by Dan Kohn »

retired at 48 wrote:Because studying one individuals performance is easily ascribed to "luck"....or a favorable investment time period, and on and on.

Most individuals near retirement did not save records "to be backtested." It is hearsay.

Academics would insist on at least 300 random cases, an impossibility to satisfy.
I think your issue is that you lack imagination on the kind of rigor that is available in evaluating market timing claims. That was the point of rereading Malkiel's chapter, not his rebuttal of any specific strategies.

We don't need any individual records, and we can test millions of different random cases in the time it takes for me to post this message. And the result of all that work has shown that there is no market timing strategy that has reliably beaten buy and hold.

But I'm not suggesting that more work can't be done. In the above post I describe the kind of rigor that might attract more attention.
retired at 48 wrote:We've heard a thousand times that nobody can beat the market, but the early retirees in my community have no choice but to accept that: "it was all luck."
R48, in every year, half of all investors beat the market (before fees). I'm not sure what your point is.
retired at 48 wrote:Lastly, I suppose you could start with this individual. His name is James Simons. He is the highest earning salary person in the world...over $1 Billion last year.

He was a professor, who felt he could develop an algorithm that could beat the market. He did. It did. He runs the most successful hedge fund on Wall Street, and has not shown Madoff tendencies. He has 200 people, daily, trading stocks. He has survived close scrutiny, returning about 24% a year for the past 15 years. Google him.

They told him it couldn't work, too, because nobody beats the market.
Now I'm really confused. Unless you are suggesting that Jim Simons has made his spectacularly great returns using a 200 DMA market timing strategy, I don't see the relevance to our conversation. (And I suspect he is doing more for his 5% management fee and 36% of profits.)

But the fact remains that hedge funds as an asset class have underperformed Boglehead-style investing, particularly on a risk-adjusted basis and after Madoff's made up returns are taken out. Yes, there have been spectacular hedge funds, just like there have been spectacular mutual funds performance. But many of them have later blown up like LTCM and Bill Miller.

The question is not whether some hedge fund, or mutual fund, or market timing strategy will beat out buy and hold in any given year. It is a certainty that one will. The point is that none of these have been able to win on a risk-adjusted basis across time. And that the capabilities of Jim Simon's 200 rocket scientists really has no bearing on the value of a 200 DMA market timing strategy.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Post by richard »

retired at 48 wrote:Lastly, I suppose you could start with this individual. His name is James Simons. He is the highest earning salary person in the world...over $1 Billion last year.
I've never really understood the argument that says because some highly talented individual has beaten the market for a number of years and made a fabulous amount of money, therefore by following a few simple rules you can beat the market on a consistent basis. I would have thought the small number of examples and the fees they command would demonstrate the opposite point.

Perhaps someone can explain what I'm missing.
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: 200 MA

Post by Dan Kohn »

retired at 48 wrote:If you mean some mutual fund can become a fund-of-funds, investing around a 200 day Moving Average (MA) scheme, perhaps it could be tried. Perhaps it has.

But an individuals portfolio, designed for his specific needs, is what is discussed throughout this thread, This cannot be lumped into mutual funds. For example, we discussed making 401.k payroll adjustments based on MA's; rebalancing around MA's instead of birthdays; changing reallocation percentages when one ages, for example, and using 200 day MA's to make those changes. Deciding one doesn't want REIT's in their allocation, and how to exit. You receive an inheritance, and have to decide how and when to invest it.

So, no, this can be done by investment advisers like a Rick Ferri, dealing with individuals, but not en mass in a common mutual fund.

There's two parts to the thread...do MA's work, and what can one practically do to utilize this tool?
Let's focus on whether 200 DMA works or not. If it does, than there will certainly be mutual funds set up that simply buy and sell SPY based on the formula you advocate. This mutual fund could then be purchased within a 401(k) or with a lump sum inheritance.

Also, please note that Rick Ferri strongly advocates against market timing and instead recommends holding bonds to reduce volatility. Rick is an editor of the new Bogleheads book, along with Taylor, and Mel, and Laura.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T »

.
My suggestion to those considering a moving average market timing approach is do your own analysis (daily data back to 1966, and monthly data back to 1927 are available on Ken French’s website), compare against alternatives (tilting to small/value, global diversification, adding bonds), and test for sensitivities reflecting personal circumstance (particularly taxes), then decide. Setting up the tests is fairly easy to do and gives a better sense of past performance and its sensitivities. Personally the results of my analysis are not as convincing as other studies seem to suggest, but perhaps I’m in the minority.

My concern with many of the backtest studies is they look at the S&P500 or TSM only, and as in the table below conclude: moving average market timing beats buy-and-hold. However, if the analysis is extended, the message seems to change. Using a 50:50 TSM:SV portfolio (which seems less susceptible to 'bubble' risk), a buy-and-hold approach 'beat' the 16 month MA (as shown by the results in the table below). In addition, the buy-and-hold portfolio had a higher annualized return in 91% of all rolling 30 yr periods. In the 9% of periods when the 16 month MA beat buy-and-hold, the average return comparison was 13.83 vs. 13.96% (all before tax considerations).

Adding exposure to additional risk factors of size and ‘value’ enhanced returns of a buy-and-hold portfolio, much more so than a moving average approach using TSM (or the factor tilted portfolio). But what about the risk reduction benefit...

Code: Select all

October 1927 to October 2008 (the longest period of data available on Ken French website)

                                    Growth of $1       Annualized SD
US TSM
...Buy-and-hold                        1,510               18.8
...16 month MA*                        2,341               12.9  

50:50 US TSM:SV**
...Buy-and-hold                       12,378               23.1
...16 month MA                         5,376               15.9

* When the ‘price’ is above 16 month moving average the portfolio is 100% equities, when it’s below the 16 month MA the portfolio is 100% t-bills. The same approach is used throughout this post for the respective equity (and equity:fixed income) portfolios.
**Has a 0.4 value load (similar to my value tilt targets), and a size load 0.45.

...If the analysis is extended to global diversification with the addition of fixed income, the stated benefits of the MA approach (enhanced return and risk reduction) seem to disappear. As with previous studies these attributes are apparent in a backtest with just TSM (second line in table below compared to the first). However when a small cap and value tilt are added, with a fixed income allocation, these stated benefits seem to disappear (compare lines 4 and 5 in the table below). Adding a global value tilt enhanced returns and lowered volatility, with no apparent additional benefit from the 16 month moving average.

Code: Select all


April 1976 to October 2008*

                                     Growth of $1       Annualized SD
                 
US TSM 
...Buy-and-hold                            29              15.3
...16 month MA                             34              12.8  

50:50 US TSM:SV*
...Buy-and-hold                            78              15.5
...16 month MA                             44              13.5
...Buy-and-hold (with 13% bonds added)     55              13.5

25:50:25 US TSM:SV:EAFE Value
...Buy-and-hold                            86              14.6
...16 month MA                             63              12.6
...Buy-and-hold (with 13% bonds added)     64              12.7

*Longest period of data available for MSCI EAFE Value (plus 16 months for the moving average analysis)

None of the above includes the impact of taxes, which would have been significantly higher for the portfolio using the moving average approach versus buy-and-hold. There are obviously no guarantees going forward, and the results are likely dependent on the extent of the value tilt and international exposure (ie. asset class), and time period used. Nevertheless, the results don't look as convincing as suggested by some of the studies (obviously just my analysis and interpretation, and no doubt many may disagree).

Personally, I will just stay the course with buy and hold, with an asset allocation that reflects targeted risk factor exposure, global diversification, and tax sensitivity (includes rebalancing and tax-loss harvesting). Will however track how the MA cross-overs correspond to a rebalancing band approach.

Robert
.
retired at 48
Posts: 1726
Joined: Tue Jan 15, 2008 6:09 pm
Location: Saratoga NY; Port Saint Lucie, FL

200 day MA

Post by retired at 48 »

Dan...You have taken a thread I didn't start, and in which several other Forum members have joined in, and personalized it to me. Yes, I am an advocate of limited use of 200 day moving averages. But if I am viewed as a personal spokesman, then OK.

But if you're going to personalize my life, what I did and didn't do, I suggest you read (or re-read) my posting titled First Time Posters IRA @ $1,000,000: Shares Wisdom Learned., here:

http://www.bogleheads.org/forum/viewtop ... sc&start=0

For that will form the basis for the following errors in your description above:

I have not done any backtesting...and the backtesting done is across a broad range of periods.

I have never used leverage in my life, or bought on margin.

My investment holdings, no stocks, all mutual funds, were perhaps held ten years in duration on average, hardly a market timer. A good portion of that was with active funds, because index funds did not exist.

I have never had a time I was below 65% equities in the market, until recently, based partly on Taylor's and Larry's books.

You state success was due to the bull market of 80's and 90's, conveniently leaving out that I went for 16 years with DJIA going from 1000 to 1000 (flat) and the period from 2000 to now, a down period.

I will close by simply stating what I posted on another thread , "What did we learn from the 2008 downturn." And that is, if I had one do-over in my investing life, it would be paying more attention, not less, to the 200 day moving average as a tool in portfolio management.

And why didn't I use it more? Because of the plethora of guru's who clouded the issue, saying one cannot time the market. I'm not accepting the smokescreen dogma anymore, and will let the data speak for itself regarding the impact on one's portfolio. Others can ponder.

But tell me, why is it Larry Swedroe can make this recent post:
SH (SmallHi)
First, I have never said that this(special tilted portfolio) is right for all investors or even most. Only those that really understand the issue and have the discipline to stay the course. Which are few and far between IMO. But if you educate people on the issue some will get it and never once have I had an issue with one who did adopt it. At least so far.


(Swedroe)
Similar disclaimers are made on this thread and any other more sophisticated thread. It's not for everyone.



retired at 48
User avatar
BlueEars
Posts: 3968
Joined: Fri Mar 09, 2007 11:15 pm
Location: West Coast

Post by BlueEars »

Robert, a few questions on your methods:
1) Are you getting TSM (1927 to present) from the FF data by using the Mkt minus Tbills data you mentioned somewhere else?

2) When you say
When the ‘price’ is above 16 month moving average the portfolio is 100% equities, when it’s below the 16 month MA the portfolio is 100% t-bills.
By price do you mean the monthly price (not current daily price)? Do you think that is an adequate trigger if it is monthly price?

3) You added enough bonds in the "25:50:25 US TSM:SV:EAFE Value" portfolio to get similiar returns and std deviation as the 16mo MA. Suppose one just wanted to use a TSM portfolio (no value tilt because you were unwilling to take the tracking error risk). How much of the bonds would be needed to make the portfolio equivalent? Could you give a few modest hints as to your method of doing this analysis? Also what did you mean by "bonds" (5yr Treasuries, TIPS, etc)?

Another somewhat unrelated question, and don't feel obligated to answer this one. You have said you use something like a 50:50 5yr_Treasury:TIPS mix for your bonds. Are you keeping this constant in light of the very low Treasury yields now? I notice that VFITX (intermediate treasury fund) has returned 14.7% over 1yr and 5.4% of that is in the last 1 month but the 10yr average is 6.2% -- perhaps approaching a low in yields.

P.S. Thanks for always keeping comments on a pleasant analytical setting. There are others on this thread that could learn a few lessons in civility :) .
Last edited by BlueEars on Mon Dec 22, 2008 10:46 pm, edited 1 time in total.
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Re: 200 day MA

Post by Dan Kohn »

retired at 48 wrote:You have taken a thread I didn't start, and in which several other Forum members have joined in, and personalized it to me. Yes, I am an advocate of limited use of 200 day moving averages. But if I am viewed as a personal spokesman, then OK.
The thread is addressed to you and you have been the chief advocate of this market timing strategy for months. You have suggested that you would like to "teach" it at the next Bogleheads meeting. Thus, I think it's fair to address you directly.

Please note, however, that I am only attacking your market timing proposal, not you personally. As I said, I believe your intentions are good in your participation here. But I believe that your ideas are going to lead to disaster for many readers, which is why I'm trying to argue against them.
retired at 48 wrote:I have not done any backtesting...and the backtesting done is across a broad range of periods.
Nor did I suggest that you have. But, R48, either take responsibility for the analysis or don't. This morning you said:
retired at 48 wrote:There comes a time when the data and backtesting stands on its own. I submit we are there.
I suspect that neither you nor I have the expertise to do an adequate job (though my wife could certainly whip up the MatLab scripts, if I asked nicely enough). But you are citing backtesting work done by others, and I'm trying to explain why what I've seen so far has been thoroughly inadequate.
retired at 48 wrote:I have never used leverage in my life, or bought on margin.
This is, I believe, not the case. You may not have used margin, but a home equity loan (HELOC) is another form of leverage. In this post in March, you said:
retired at 48 wrote:After home purchase, take out max HELOC possible, for two reasons. First,, like you said, it is a good emergency fund backup. Secondly, use it to fund your IRA's to the fullest. This may seem controversial, but will greatly payoff in long run....
I would further not invest IRA's, now, in bonds but rather an asset allocated equity portfolio. Then even at the low end of 7% return long term one is way ahead of HELOC's. I did this my whole career. In fact I have a large HELOC today, using it for tax space to convert my regular IRA's to Roth IRA's.
I think my arguments from March against using leverage (HELOCs) to buy equities look rather prescient 9 months later:
Dan Kohn wrote:retired at 48, you're suggesting that young investors use leverage to increase their returns. This works great, until it doesn't. Leverage increases your returns in good times, and creates the potential for bankruptcy in bad.

You don't need to use a HELOC; you can just buy stocks on margin. For the downside of that experience, read about our own market timer.
http://www.diehards.org/forum/viewtopic.php?t=11742

If you think it can't happen with home equity, you're wrong. We're facing a recession where your home could fall 20% in value, the market goes down, and you lose your job (even teachers). This means you could face a real risk of losing your home and having to declare bankruptcy, just so that you could try to juice your retirement returns.

And yes, retired at 48, I appreciate that leverage worked for you and that you were able to retire early. But your prime earning years also happened to coincide with the longest bull market in the history of the world. For an alternative view of the potential of leverage, take a look at Capital Decimation Partners:
http://krugman.blogs.nytimes.com/2008/0 ... -partners/

Admittedly, CDP has negative expected returns while your proposal probably has positive ones, but it's the tail of the distribution that kills you, not the mean.

I would strongly recommend against leverage except when used for a house. Even then, I would strongly encourage a 20% down payment.
retired at 48 wrote:My investment holdings, no stocks, all mutual funds, were perhaps held ten years in duration on average, hardly a market timer.... I will close by simply stating what I posted on another thread , "What did we learn from the 2008 downturn." And that is, if I had one do-over in my investing life, it would be paying more attention, not less, to the 200 day moving average as a tool in portfolio management.
Either you're advocating for 200 DMA or you're not. People can't "sort of" market time. The fact that you may not have done as much market timing when you were saving money as you are promoting now doesn't impact the reliability of your system.

I certainly hope your advice to new investors who have been burned by the market is not to evaluate the 200 DMA market timing signals and then to make "gut feel" decisions about whether to get back into the market or not. You've made hundreds of posts espousing a market timing system. Why stop now?
retired at 48 wrote:Similar disclaimers are made on this thread and any other more sophisticated thread. It's not for everyone.
That seems like a ridiculous cop out to me. If a 200 DMA strategy reliably beats buy and hold, why would you possibly suggest that every investor shouldn't use it? I've already shown that (whether it works or not) you could easily package the strategy up into a mutual fund that anyone's grandmother could buy. Why would you keep this luscious, forbidden fruit from investors who just want the same returns at lower risk?
SP-diceman
Posts: 3968
Joined: Sun Oct 05, 2008 9:17 am

Post by SP-diceman »

Dan Kohn wrote:
As (I hope) you know, this is an absolutely ludicrous example.
Its not a ludicrous example.

You just don't like the facts.

(the high is irrelevant. If they were equal 7%
before the high. Then they will be equal at
the high)

The only point is that they are relativity equal.
Would you like me to analyze the B/H investor
with $30,000 and the MAV with $500.
Would that be a "fair" comparison?

I probably "short-changed" the market timer by
making a huge run up to the buy area to illustrate
the out come.
(remember the larger the run-up the more expensive
the buy-back. The more the B/H portfolio recovers)

The reality is each day there is no buy the pendulum
swings more in favor of the timer.
If the market stays flat or down. Shares will
only get cheaper and the MAV will get lower.

Making the buy happen moved the B/H portfolio to
$7697 and buying in at 121.26.
If we don't have the run-up the B/H wont reach that level
and the MAV will only buy cheaper.


Thanks
SP-diceman
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Examples

Post by Dan Kohn »

SP-diceman wrote:Its not a ludicrous example.
What I'm saying is that if the period you test data on starts at a peak and ends at a trough, than any market timing strategy that gets you out of the market will beat a buy and hold strategy.

When the average is up over time, buy and hold will generally outperform, sometimes by enormous amounts. To make any interesting statements about the value of a mechanical trading technique, you need to test it across a large number of different periods.

Here is a nice review of some mechanical trading patterns that were recommended in the past, because they had performed well on data mining. They all failed on a going forward basis.

From Investor Home
http://www.investorhome.com/mining.htm

And Bernstein's summary of this material
http://news.morningstar.com/articlenet/ ... spx?id=568
Speedie
Posts: 295
Joined: Thu Aug 23, 2007 6:52 am

Post by Speedie »

Robert T wrote:.
Data for the longest period I have (data excludes tax impacts).

Code: Select all

US SMALL VALUE - 1927/10 to 2008/11

                             Annualized Return     Growth $1

Buy and hold                      14.48              57,610
Market timing (16 month MA)       13.26              24,272

Source: Small value series from the FF 2x3 file on Ken French’s website

This is why I’m also cautious of any sweeping statements on the superiority of marketing timing over buy-and-hold. Backtest results are sensitive to asset class used, time period, duration of moving averages, buy and sell bands around the MA, and taxes (perhaps more). Of all the marketing timing back-tests I have done, the results personally don’t change my top four asset allocation priorities (for better or worse):
  • 1. Diversify across risk factors (market, size, value, term)
    2. Diversify globally (US, Non-US, EM)
    3. Minimize costs (including taxes)
    4. Rebalance (using rebalancing bands).
I will continue to track performance of some MA approaches, which may inform rebalancing frequency, but the more I look at the data, the more doubtful I become (particularly for a value and small cap tilted portfolio). Time will tell.

Just my personal view.

Robert
.
One question Robert: did you make any allowance for interest earned when the 16 month MA was in cash? I suspect that it would have been for a number of lengthy periods, and if interest was earned while out of the market (surely a realistic proposition) then the system return would have been considerably higher due to compounding.
idoc2020
Posts: 1090
Joined: Mon Oct 22, 2007 3:40 pm

Post by idoc2020 »

Dan Kohn wrote:
What you need to understand is that the 200 DMA strategy (and, indeed, any market timing strategy) can kill you on both the upside and the downside. It's not hard to imagine a new bear market that completely slaughters those using this strategy. All it would take is a really drastic fall that occurred while you were in the market (such as from a horrible news event). Then, you could have a short number of drastic increases followed by slow declines, such that you were always out of the market when the big increases came.

The fact that this last bear market would have been avoided by 200 DMA tells you nothing. That's what the complaints about data mining are trying to communicate. Because there is no guarantee whatsoever that the next bear market will look anything like this. That's what a random walk means.
Depends how you use this strategy. We have already concluded that a drastic bear market such as occurred in 1987 would bypass this strategy due to its speed. However, the MA timing concept could be used in other ways eg: to figure out when to get out of a market at the tail end of a bullish period. Or when to get into the market at the tail end of a bearish period. Or when to lump sum or DCA incrementally into the market. Or perhaps when to rotate your assets. So yes, there will be market moves that may preclude adequate utilization of this strategy but it does not obviate its utility.
User avatar
Dan Kohn
Posts: 1505
Joined: Tue Jun 26, 2007 12:15 am
Location: New York, NY
Contact:

Formatting request

Post by Dan Kohn »

When you reply to someone else's post on this forum, please click the quote button at the top right of the post. That sets up a reply with the previous text fully quoted, and makes it very clear whose text is whose. If you're curious how it will look, click Preview before you hit Submit and you can confirm that the quoting (or double-quoting) is working correctly. Thanks.
Post Reply