If the font is too small you can hit Cntrl+ to make it larger (this works for Firefox and IE; hat tip to an anonymous reader) Cntrl- to reduce.I read yet another article yesterday that said the market cannot be timed. This is a good rule of thumb to be sure but is not universal and for anyone who is new it might be worth giving one example of what has been a useful tool and what I use as a tipping point in the accounts I manage.
The first chart is from the mid-1970s and the second chart is from early in this decade. The black line in both charts is the S&P 500 and the yellowish or brown line is the 200 DMA.As you can see in both instances (and it is also true in the current market event) the 200 DMA was not breached at the top but served to avoid most of the decline if adhered to in the strictest sense (which would be all in above and all out below).
If you look at the S&P with its 200 DMA over the last few years you will see several headfakes
with this indicator where it went below the 200 DMA for a few days before going back above. This would not matter to anyone with one ETF. A fake out that results in one or two extra $10 commissions is not a big problem.
However in the real world where investors have diversified portfolios that would incur real commission expense and tax consequences all in or all out becomes impractical. But it is the evidence in the two charts (and how the current decline has played out is further support) which leads me to use this indicator for taking the type of defensive action I have mentioned numerous times before.
Changing subjects... the Fed seems to have pulled a rabbit out of its hat yesterday with the TSLF news to help lube the bond market's chassis a bit (although some question what the real catalyst was) and despite my bearish expectations I would be thrilled if the bottom was in and equity prices would stabilize with an upward bias (pretty much Emily Litella-izing yesterday's post). Bull markets make the job easier and more importantly makes it easier for clients.
You probably heard the comments from the various naysayers about the TSLF and unfortunately I probably side with them. There are still more writedowns to come, I can't see how earnings don't get further impacted, one reader forwarded a skeptical report from Merrill that cited the market is $6 trillion big and the TSLF is only $200 billion, bear markets usually turn for no reason at all not from a clear and obvious catalyst and there are probably other things too.
One other aspect about this is that the Fed is clearly willing to take action that appears to be innovative and has no qualms about doing something and then coming back very quickly with something else. While that could be a plus it leaves me thinking that the Fed is reacting to news as it hits the tape instead proactively trying to right the ship. If that is correct then I don't take much solace from a Fed that is continually a step or two behind the music.
Here's hoping I am wrong.
31 comments:
Roger,
While lacking the techno skills to reproduce the chart here, I note that in a 2 year chart of $SPX using the 200 dma marked against the 50dma, (taking crossovers as significant signals), it is worth noticing that there was only one head fake (lasting about two months) in July of 2006, but otherwise, the signals were: Short (headfake) at 1242 (close of July 13, 2006), Long at 1316 (close of Sept 13, 2006) and Short, 1484 (close of Dec 21, 2007). The 50dma is still way below the 200dma.
As you've mentioned in past posts, one thing that is distinctive about this period: the 200 DMA has a negative slope since something like January 15 of this year.
These two measures (200dma and 50dma) are not "predicting" anything - but as you've repeatedly pointed out, they do reflect the deeper sentiment of the market as a whole, and, as the above crossovers seem to indicate, offer some cooler distant perspective during what can be quite volatile daily and even weekly gyrations. (See, for example, the 180 pt swings on $SPX between July 16, August 15 and Oct 10 of this year: the 50dma never touched the 200 dma.)
R in NY
Timing naysayers don't understand the goal of market timing - risk reduction, not enhanced performance. I agree with Paul Merriman's view of timing as presented here:
http://www.fundadvice.com/fehtml/mtstrategies/0104.html
The example Merriman uses is the SPX vs. the 100 day moving average from 1942-2000. Note: the 100 day MA has no special magic - use 75, 150, 200...whatever.
Here are the annualized returns (%) 1942-2000:
SPX: 13.3
Timing: 12.7
Timing with 1.5 daily leverage: 18.3
A simpleton would conclude timing doesn't work. But lets look at the standard deviation:
Standard Deviation (%):
SPX 16.2
Timing 12.4
Timing x1.5 daily: 19.9
Hmmmmmm. Wouldn't you give up .6% a year for a lot less volatility? Or even use a combination of timing and some leverage?
Timing isn't for most people because it requires tremendous discipline to execute, but people who say timing doesn't work are using the wrong yardstick.
Roger:
What about all the new money being pumped into the system to ease the current financial problems? Is'nt that going to lead to one awful bout of inflation afterwards? Should not a good defense be to have investments that will respond to this extra stimulus? Any ideas of sectors that usually are most affected by inflation?
Willy
Willy,
i have touched on that many times.
Here is a link from me and one from Mike Shedlock that address the question.
Agree that individual investors must focus on the risk reduction. Decades ago when Stockmarket Logic was first published this was the second part of Fosback's argument: If you were only accepting market risk part of the time then virtually by definition you're risk-adjusted returns were higher.
The first part of Fosback's argument back then was that you could time the market for enhanced performance too but I think in these days of large-scale, program trading that has become increasingly difficult: Not impossible necessarily but the discipline and level of knowledge required are probably much higher than most individual investors can achieve.
In that vein there's an interesting paper being discussed at Mark Thoma's site (http://tinyurl.com/yrtbjx) the upshot being as the title of the paper states that highly predictable volume leads to highly predictable returns. Baldly stated it would seem that if you can move a market you can time it, if you can't then good luck.
And good luck too trying to follow the timers because it's pretty much all black box and proprietary (and inside) info and heavy leverage now so the individual investor who attempts to emulate will either miss the timing or misinterpret the real play or have inadequate leverage or all of the above. JMO
Read "A Random Walk Down Wall Street", or Swenson's Pioneer book. There is no reason to believe such models work, b/c in the past they never have. They are right until they are wrong. And if you rely on them heavily the results can be disaster. I understand Eckhardt's opinion on the matter, and perhaps you can find needles in the haystack, but I don't think betting on the whole haystack is useful.
I take the safer strategy of buying when values seem better than they have been recently (i.e. pullbacks). Hardly foolproof, but it beats dollar cost averaging.
"There is no reason to believe such models work, b/c in the past they never have."
The two charts on the post would seem to say otherwise.
Those interested in timing the market (or other trading systems) might enjoy the evaluations at cxoadvisory.com/blog/. They find that very few systems consistently beat buy-and-hold and those that do are usually momentum based. That's probably an over-generalization, but that's my take.
Fine, use technical analysis. How do I go short your firm?
(I personally don't care what you do, and I'm not going to bother summarizing the points described by Malkiel and re-emphasized by Swensen -- I will simply say two things:
1. There exists no known correlation between past prices and future prices that has held over time -- particularly at the aggregate -- I wish there were.
2. Why is it the 200 dma, and not the 201 dma, or the 187 dma? Your system is arbitrary, price variance may not be normal, but it is apparently random, sad to say.)
so are you saying there is only one way to do things?
that being the way that makes the most sense to you?
BTW 200 DMA as an indicator of relevance is hardly a unique thing that I came up with. Plenty of people use it.
BTW, if you have Vista and Ctrl+ doesn't work on your browser you can go to your toolbar in the upper right hand side of your page and click on 'page'. Then click on 'zoom' to increase or decrease the page and font size.
And good info on market timing Roger. Too many investors still don't believe that defensive measures are necessary in their portfolio.
I just read a post on Market Watch where a 54 year old man with "a 6-10 year horizon" doesn't need to time the market going into a bear market. He must want to work until he's 90 I suppose.
54--6-10 years?
that might be his horizon until retirement but no one should have a 6-10 time horizon.
someone who is 110 with millions has a longer time horizon than 6-10 years (that being for whomever the money would be left to; family or charity).
No the point is the 200dma is arbitrary, i.e. there is no reason why it is any better than a 199dma or a 211dma. Your results using it will also be arbitrary. That is, there is no reason to believe it will beat the market, though it may. OTOH, Eckhardt (a technical trader) and Swensen (a fundamentalist) have beat the market for years, and yet they have explicitly rejected your method. I am unaware of anyone with a long *successful* track record who approves of your method. But hey, maybe you're 200dma is really cutting edge, and we will all be reading your book's, rather than David Swensen's in 10 years.
quite the opposite of cutting edge.
i have made the same point in past posts that it could just as easily be 195 or 205.
you can say all you want to poo-pooh it but the charts posted along with the current cycle say otherwise.
from a forest from the trees standpoint just look at the charts and also think about what is really happening to cause the market to go below its 198 DMA.
one more thing as far as NOT cutting edge,
i first read about this in what I believe was an article written by James Stack in Smart Money Magazine in 1993 or 1994 and I doubt he made it up then, so nothing cutting edge about it which is exactly the point.
I don't know anything about your "200 DMA" voodoo, but I do now that an original NFL team won the Super Bowl this year, so back-up the truck! LOL
Seriously though, I do like moving averages because they tend to remove all ambiguity surrounding an assets trend. Show a six year old a chart of the 200 DMA (or 190 or 205 - it's irrelevant), and they can instantly tell you if the asset is going up, or going down. In fact, my youngest daughter told me to buy TWM this morning. She's such a sweety!
Just downloaded the daily prices for the UK's FT100 April 1984 to present from yahoo finance.
Testing for where the the current days price was greater than the 200 dma and taking the days index gain factor where so, or using 1.0 gain factor where not results in an overall 1.88 gain factor versus 4.48 for the Index. Only has 70% in-time however, so scaling down the index to the same exposure time averaged 2.85. Still no better.
Being in only when the current was > 200 dma averaged a worse result.
Taking the Dow however, 1928 to present had a 124 gain factor versus 34 for buy-and-hold, and with only 65% in-time.
Reducing the Dow sample down to 1990 to present however saw 1.98 versus 4.26 (74% in time).
So over the very long term the 200 dma timing method would appear to have worked, but as of 1990 to present (or 1984 for the UK) it hasn't.
And bear in mind ;>) that these figures exclude the potentially quite significant cost of switching in and out.
and for those who lack the market-timing gene:
a simple re-balancing every now and then accomplishes at least some of the desired effect.
assuming that one has CASH as one of the allocations.
some people just go "all in" equities, for reasons which escape me.
Clive, thanks for the leg work, I did mention that in a one index fund portfolio it would be easy to execute but that it would not be practical for a normal diversified portfolio (all in versus all out) but as a means to begin some defensive action it stand up for me unambiguously; the two charts and my track record as anecdotes to support the theory.
So despite the fact that you will do worse than simply buying an index fund and doing nothing (at least based on holding for only the last 18yrs vs the last 80), you will continue on your 200dma strategy (and really those time ranges are as arbitrary as the 200 you've chosen).
in addition to Clive saying it works for the US, you can check my quarterly performance recaps to see what my results have been.
Thanks Roger.
I ran yet another test for the SPX over the Jan 2000 to Jan 2004 period indicated in your graph and found that yes in that case the 200dma did provide better results than buy-and-hold.
36% in time in total, 7% loss versus 24.5% loss for buy-and-hold.
But sticking to rotating in and out as the price moved above/below the 200ma based on daily open prices incurred 35 rotation trades. Potentially quite costly subject of course to how large the funds were that were being managed under the style.
So worked as a possible indicator over Dow 1928 to 2008, but not over Dow 1990 to 2008. Didn't work for the UK either 1984 to 2008. But worked for the SPX 2000 to 2004.
My guess FWIW is that its a potential indicator of troubled times (as you say begin some defensive action) that if such comes to fruition avoids potentially large losses.
I've observed many times before how many who claim timing doesn't work usually cite that you miss the small number of days that provide something like 90% (or whatever) of the upside gains - but usually ignore the counter Bell Curve side of how missing the small number of days having the largest losses counters the argument.
A similar argument might equally be made for running with stops in many respects.
Those most troubled and largest losses don't usually come out of the blue (excepting the likes of 911).
Clive you isolated a crucial point in that the data can be spun (mined) to show anything.
so the most important thing is doing what is comfortable for your own temperament. for some people that means taking little to no action at one extreme to constant action at the other extreme. Both can work and both can fail so hopefully people recognize when they are failing but there can be so single answer that is right for everyone.
I opine that timing, as in rebalancing, does add value - in the form of potentially comparable (or better) rewards for less risk.
If you take a simple case of yearly rebalancing back to cash/stock equal capital weights. Measure the rate of return against stock value and the (XIRR) will generally be higher than for buy-and-hold.
Cash however acts as a drag on the overall performance.
Typically simple yearly rebalance will uplift stock XIRR by 0.5% or so. Improve the dynamics of rebalancing however in a variable/timely manner and that stock value XIRR can be uplifted by 1.5% (I personally use Robert Lichello's AIM for such timing).
On that basis given cash (bond) returns of 8%, buy-and-hold returns of 12% coupled with a 1.5% uplift in stock value XIRR then 8x + 13.5y = 12 solves (as x+y=1) y=0.727 which means that given 72.7% average stock exposure at 13.5% and 27.3% bond exposure at 8% then the combination paces 100% buy-and-hold at a 12% return. Too little average stock exposure however and returns underperform buy-and-hold.
Yet a further benefit is that a dynamic stock/bond blend typically incurs lower volatility than buy-and-hold and as such the compound benefits over time tend closer to the (higher) simple average.
It gets even more interesting when you rebalance across multiple levels ;>)
I like to rebalance at the top-most (market wide stock/bond allocations) level, within individual accounts (AIM individual stock holdings), between AIM accounts (effectively between sectors) and via a virtual bond holding (stock/cash blend). In my experience this has the effect of moving the line significantly downwards where the level of average stock exposure required to pace buy and hold occurs, whilst achieving such in a highly mechanical/automatic manner. You just have to follow the signals. A significant benefit, at least for me, is that the more faith you build in the signals over time the more you're likely to follow the mechanical signals rather than over-riding with (often wrong) human emotional trades based on (often wrong) natural human patterns seen in waveforms.
Works for me at least (single parent dad with no other form of income other than that from my investments).
Wow, looks like a few folks located random's hot button. This 200 DMA on the SPY looks handy, i think i'll combine it with the MACD to help me with adjustments in my 401K. Useful post random, thanks.
Certainly is a lot of energy and interest focused on the question of "is it better to time the market than buy and hold"?
Couple of thoughts:
If you EVER sell, you have to decide WHEN to sell. On the roughest measure, you do "time" the sale (unless you're forced to liquidate - cf. Carlyle, Soc Gen, etc. ;-). While that's not the intended meaning of "timing the market", it does point up that exit decisions are difficult, and it may be a relief to focus on data that supports the conclusion that "now's as good a time as any".
But, if you concede that some thought/analysis may be worthwhile as to when to exit (or enter) a position (after/before an earnings announcement, after/before an dividend announcement, after before economic announcement, after/before whatever you think may adversely/positively impact the price, (and I grant that "strong" market theorists won't concede this point and will instead argue - quite wrongly in my view - that the price reflects all information, but whatever, then it is a short step to considering a variety of factors in your analysis. Including whether the market (for that sector, for all sectors, for that industry, for that stock) is performing well, or is not.
I think all that Roger has pointed out is that the 200 dma apparently does offer some information regarding the strength of the SP500 as a whole.
But here's the second, and to me, more interesting point. The 200 dma, like the old mid-1980s days of Joe Granville announcements, may be somewhat self-fulfilling. For example, I consider the 50dma crossing the 200dma something of a self-fulfilling indicator. When traders are all looking to [something like the same thing - whether it is the 200 dma or the 205 day, or 197 day], and watching the 50dma (or something like that), and you see price breaks from that level (at the crossing), it really doesn't matter what is happening fundamentally - sentiment, as we're seeing in spades these days, can drive prices.
While much of the discussion has been, and rightly is, "investor" based (which I understand to be trading based, but with a longer holding period), I don't think the distinction merits dismissing the possible benefits from including the [limited, somewhat unreliable?, theoretically not supported] information from market psychology signals like the 200 dma.
No one says you have to use it. For so long as many people do use it, and the market expresses sentiment at certain commonly observed "coincidences", it may be a useful tool to consider using.
Rick in NY
"someone who is 110 with millions has a longer time horizon than 6-10 years (that being for whomever the money would be left to; family or charity)."
Unfortunately, most people do not
think like you Roger. I still
remember a talk with a man who
said life would be wonderful if
he just knew when he was going
to die....so he could spend every
last dime. There are also many
spouses who tell their spouse
to "sign off" on their pension...
(so they collect 100%)
and they will get life insurance
instead...only to find out that
the spouse canceled the policy
a year later...and are left with
nothing but hospital bills after
they die.
http://www.smartmoney.com/barrons/index.cfm?story=20080312-are-you-rich
We have lots of work to do;-)
Roger,
The Dow v Gold issue is creeping up again.
http://home.earthlink.net/~intelligentbear/com-dow-au.htm
In 2004, you wrote a little about an article in Barrons, where the gold bug talked about some future credit issue, blah, blah...
Well, I would like your take on this now. The goldies say that the ratio gets back to "1"....so the DJIA goes down and or the Gold goes up....what do you think?
g
George, maybe i'll address this in the next video but...
I am by nature not pessimistic enough to think a ratio of 1 to 1 is possible so I may be the wrong person to ask.
What sort of price might they meet at? $3000? $4000.
Can gold triple or quadruple after having previously quadrupled?
Can the Dow realistically go from 14000 down to 3000 or 4000? Dow 4000 would be a 70% decline. Japan went down a little more. My perception is that Japan is/was set up in such a way that its flaws are different from the US flaws. Part of Japan's problems as i understand is they did not write down assets that should have been written down, the US has had a lot write downs with more on the way.
The US' flaws seem to create excesses that we then have to pay for.
Of course I may have the whole thing upside down.
Roger - You should overlay the 50 week DMA and the 20 week DMA on a weekly chart. Historically over the last 10 years or so, when they cross over by 1% or so, then you know that the bull or bear has run it's course. Obviously we are in a different market altogether now, but if you chart this, you can see that the big money appears to use this measure as well. We've got a ways to go in this bear, it appears, at least based on divergence in this measure.
Post a Comment