Momentum - who do you love.

[TickTock]

Well, the mutual fund that they put together on this has underperformed it's index (at least the one that Yahoo Finance chose) in 6 out of 8 of the time periods that they report.

How will that help me?

FUNDX: Performance for FUND X UPGRADER FUND - Yahoo! Finance

-ERD50

Yahoo grades them against large blend international. I don't know how fair a comparison that is (literally don't know - fundx upgrader currently holds 55 funds! I saw a bunch of names that clearly said international, others domestic, others I can't place from the name) - does fundx change allocations in their asset classes, or only in the funds within those asset classes? I couldn't find it on their website.

 

[brewer12345]

I tend to swing the way ha does (don't wanna own it if I can't understand and get comfy with the business), but this is an interesting topis nonetheless. I get the feeling that the serious academic literature on the subject would quickly be over my head on the statistics end of things.

 

[W2R]

W2r,

Thanks for all the quotes!

You're welcome - - glad you liked them!

I just hope that by limiting the quotes to a sentence here, a brief paragraph there, and by giving the citations and page numbers, my attributions to the sources were sufficient to meet any definition of limited use. I have read the same thing in every book I have, but these quotes were from the first four that I happened to have at hand. I make no secret of the fact that I am still learning so this gives some sources of what I feel I am learning.

 

[W2R]

I tend to swing the way ha does (don't wanna own it if I can't understand and get comfy with the business), but this is an interesting topis nonetheless. I get the feeling that the serious academic literature on the subject would quickly be over my head on the statistics end of things.

I seriously doubt it! You don't impress me as being mathematically illiterate, though it might take some time for any of us to delve deeply into the relevant statistical analyses.

I agree with Ha too, especially since so many sources vehemently assert that persistence can be a really, really bad strategy to follow. It's enough to make a beginning investor sit up and take notice.

 

[brewer12345]

I seriously doubt it! You don't impress me as being mathematically illiterate,

Appearances can be deceiving. I usually have to call DW when I need to do algebra. Don't ask about calculus.

You can do most any amount of accounting with 5th grade math. Finance might require high school math for most of it until you get to the seriously quantitative stuff or econometrics. I was recruited for a finance PhD, but declined when I realized how many years I would spend taking math and stats classes.

 

[W2R]

Appearances can be deceiving. I usually have to call DW when I need to do algebra. Don't ask about calculus.

You can do most any amount of accounting with 5th grade math. Finance might require high school math for most of it until you get to the seriously quantitative stuff or econometrics. I was recruited for a finance PhD, but declined when I realized how many years I would spend taking math and stats classes.

Gee - - I admit that I am amazed.

Somebody should have let you in on the secret that math is fun, especially advanced math. Math gets a bad rap due to the tedious rote memorization of multiplication tables and such in elementary school, I guess.

 

[DJRR]

I agree with Ha too, especially since so many sources vehemently assert that persistence can be a really, really bad strategy to follow. It's enough to make a beginning investor sit up and take notice.

As Twaddle said I think it goes back to your definition of momentum/persistance. My first pass at a definition is:

Options:

1. Buy and hold with a fixed allocation, rebalance, etc. - Standard type of approach for most on the board

2. Cramer - hyperactive trading high costs, taxes, etc. Chase returns, buy high, sell low, etc. Works for market makers and hedge fund types but probably not wise for the average joe.

3. Momentum investing - fixed allocation and rebalancing, but instead of holding funds forever you rotate them based upon who is performing best at the present time.

Option 2 is what everyone says you shouldn't do.

Option 3 is what I try to follow. Some further comments:

Option 1 says that you can never know what investment will perform best in the future. Option 3 says you can - funds that have momentum maintain that momentum (persist) into the short term future.

Active managers have a hard time changing styles. A deep value manager will underperform during the late 90's and overperform in the early 00's. If you hold for the whole time you will do OK, but why not sell him in the 90's and buy him in the 00's depending on how his relative performance is doing.

How you determine momentum is an open question. My method ignores 1 mo results because that greatly increase the churn, but No Load X has been successful using the 1 mo momentum results.

I would not do this with individual stocks -too much specific risk - the market could gap down, etc., but it seems to work well with mutual funds.

The strategy underperforms when there is no clear trend. This tends to churn when the market is trying to decide where it is going.

The strategy outperforms when there is a clear trend and it continues. For ex., the outperformance of small cap value was huge in the 00's and continued for many years. I got into RS partners after a large increase, but rode it for another 120% gain.

Part of the outperformance seems to be from allocation creep. For example the best performing large cap mutual fund managers moved into the mid cap space in the 00's. Therefore I was really more heavily allocated towards mid/small cap than planned.

Academic results are mixed, but tend to say that any advantage is eaten by taxes and costs. IRA eliminates the tax issue, but transaction costs can indeed be higher.

I will try to post more thoughts, but encourage you to respond.

 

[DJRR]

What you are saying is true if you are adding to your index portfolio, but if you are just holding it there is no adjusting (buying or selling) going on within the index. Cap weighting takes care of itself.

I disagree with this. Even if you just hold, the manager of the index must buy more of the increasing stocks in order to track the index. Think of homebuilders. In 2000 the S&P had a low value for homebuilders but in 2005 it had a very high value for homebuilders. Even if you added no money your exposure to homebuilders increased dramatically. You are now experiencing a large reduction in homebuilders.

 

[DJRR]

Appearances can be deceiving. I usually have to call DW when I need to do algebra. Don't ask about calculus.

You can do most any amount of accounting with 5th grade math. Finance might require high school math for most of it until you get to the seriously quantitative stuff or econometrics. I was recruited for a finance PhD, but declined when I realized how many years I would spend taking math and stats classes.

I resemble that I thought about a PHD program also, but got bored with Masters level statistics and don't even talk about linear algebra

At least Excel does most of the heavy lifting these days

 

[brewer12345]

At least Excel does most of the heavy lifting these days

And you should see what a Bloomberg terminal can do.

 

[ERD50]

I disagree with this. Even if you just hold, the manager of the index must buy more of the increasing stocks in order to track the index.

I don't think so. In fact, I'm pretty sure that is why cap-weighted funds are offered - it takes no action on the 'managers' part to maintain the weighting.

Take a simple example of an imaginary index fund with two equal priced stocks, A and B. The fund owns 100 shares of each at $100/share, so $10,000 of each stock. The fund is worth $20,000. Each stock represents 50% of the cap weighting of the fund.

Now, stock A doubles in price. The fund is now worth $30,000; $20,000 of A, $10,000 of B. So, stock A represents 2/3 of the fund, which is what it's cap weighting indicates it should, and the 'manager' did no buying or selling to achieve this.

Now, if $30,000 of 'new money' came into the fund, the manager would buy 100 shares of each again, and the cap weighting would still be in balance. No 'brain' required of the manager. True, he did buy $20,000 of stock A, and $10,000 of stock B.

A Fund's public performance is not measured by any weighting of how much money they managed, it is a straight calculation of beginning and ending NAV and dividends. If they rose 5% over 11 months, then got an infusion of deposits on day 1 of the 12th month which doubled their size, and then rose 10% in that last month, their performance would be 15.5% (1.05*1.1). It would not be weighted towards that 10% in the last month, just beginning NAV and ending NAV. Right?

Also....

Good post on the options of the momentum players. One thing that sometimes gets missed when comparing performance is the risk-adjusted performance. I'd be interested to see anything that shows a momentum fund can beat the index with less risk.

-ERD50

 

[TickTock]

DJRR,

Thanks for the post on momentum investing! So AA is fixed and only the funds within are varied.

 

[TickTock]

The strategy underperforms when there is no clear trend. This tends to churn when the market is trying to decide where it is going.

The strategy outperforms when there is a clear trend and it continues. For ex., the outperformance of small cap value was huge in the 00's and continued for many years. I got into RS partners after a large increase, but rode it for another 120% gain.

Thinking about this some more... wouldn't the strategy also underperform when there is a clear down trend?

 

[DJRR]

DJRR,

Thanks for the post on momentum investing! So AA is fixed and only the funds within are varied.

The version I use and find value in at least. Momentum per se is probably more known as using a relative strength forumula as per Investors Business Daily.

The method I use is a follows:

1. Pick your split of stock, cash and bonds - I currently use 100% stock

2. Split your stock portion into 5 buckets

1. Large Cap Value
2. Large Cap Growth
3. Small Cap Value
4. Small Cap Growth
5. International

These %'s are fixed for the year and you can rebalance as you desire. I rebalance at least once per year, and since I am selling at various times during the year as I replace funds I take that opportunity to rebalance.

Then within each category I select the current recommended funds based on their momentum rankings. These are kept until they fall below the top quartile for all funds in that category. If they fall out they are sold and replaced with new funds.

I personalize it by only using 10 funds - 2 in each category and by choosing the smallest funds by asset size because these usually exhibit greater outperformance.

That is what I do in a nutshell.

 

[TickTock]

Thanks for the further description!

I called fundx - they do NOT use AA and rebalancing; they use their "Class 1, 2, 3, 4, 5" system, and while they have base levels of each category they do allow the fund classification percentages to vary.

Therefore, they can and do vary their AA. Given this, it's not appropriate to rate them against large blend international for the past five years (unless the bulk of the fund has been in LBI for that period of time. Determining that is more work than I want to do at this point).

That brings up another point - how do you properly rate a fund that varies its AA? Something I've wondered from time to time...

The fundx upgrader fund's objective is to provide similar volatility to and better returns than large US stocks.

 

[DJRR]

Thinking about this some more... wouldn't the strategy also underperform when there is a clear down trend?

Since it is a "relative" stategy it will still outperfom if the trend is strong. You will still lose money, but lose less than the market.

 

[haha]

Since it is a "relative" stategy it will still outperfom if the trend is strong. You will still lose money, but lose less than the market.

I think I understand, and you have had good luck with this method. But this situation occurs to me- the market is uptrending, you wind up in an outperforming fund, which is almost guaranteed by the nature of things to also have a higher than market beta.

Then the market turns. For while, before your method puts you into another fund, you are going down with a that higher beta, thus losing more than the market. Eventually you 'll be in a lower beta situation. Say the market reverses again- you are now going up with the lower beta fund, and therefore losing to the benchmark.

Isn't this just a a more complicated manifestation of the usual whipsaw problem with garden variety market timing in choppy, non-trending markets?

Also, a side question-do you ever wind up in funds that are negative beta? Like short, or ultra short funds?

Ha

 

[ERD50]

Well, that was my experience. If FUNDX is a reasonable proxy for this strategy, it seems to dip pretty steeply:

FUNDX: Basic Chart for FUND X UPGRADER FUND - Yahoo! Finance

-ERD50

 

[DJRR]

I think I understand, and you have had good luck with this method. But this situation occurs to me- the market is uptrending, you wind up in an outperforming fund, which is almost guaranteed by the tnature of things to also have a higher than market beta.

I believe this is true. Higher beta is a result of higher volatility than the tracked index. I want higher beta on the upside What is unclear is whether this higher beta is a result of higher risk or superior skill or luck, ie was Peter Lynch just lucky or really more skilled. The risk is not necessarily higher unless your paradigm requires that.

Then the market turns. For while, before your method puts you into another fund, you are going down with a that higher beta, thus losing more than the market. Eventually you 'll be in a lower beta situation. Say the market reverses again- you are now going up with the lower beta fund, and therefore losing to the benchmark.

My experience is that this can be true. It is important to follow the mechanical rule and sell immediately when the momentum breaks. I have "saved" a $75 commission and lost $3,000 because a fund cratered when I held too long. Also, during market turning points you can get whipsawed until a new trend is established. Since the bias of the market trends up the gains tend to outpace the losses, but you will experience periods of underperformance.

Isn't this just a a more complicated manifestation of the usual whipsaw problem with garden variety market timing in choppy, non-trending markets?

I don't believe so because I am always 100% invested in the market. As they say, time in the market is more important than timing the market. I am always fully invested, just trying to be in the best assets within each asset class.

Also, a side question-do you ever wind up in funds that are negative beta? Like short, or ultra short funds?

Ha

No. I practice a long only strategy, which is another reason why it is different from traditional market timing. That is also why you will certainly experience losses in a bear market. The newsletter I follow was down by 14.2% in 2002. The Wilshire 5000 was down 20% so it was a relatively lower loss, but still a significant one. I believe that the upward bias in stock growth makes shorting a long term losers game and I do not attempt to time in that way.

 

[DJRR]

Well, that was my experience. If FUNDX is a reasonable proxy for this strategy, it seems to dip pretty steeply:

FUNDX: Basic Chart for FUND X UPGRADER FUND - Yahoo! Finance

-ERD50

Yes, I think this is relective of my own results:

YTD: -9.7%
1 year +15.10%
3 year +16.64%
5 year +19.05%
Beta 1.23
Standard Dev 11.99%

And remember this is after paying an additional 2.17% fee on top of the underlying mutual fund fees.

Traditional theory would say they should be underperforming by the fee, and yet they do not, and you can duplicate the strategy yourself without paying the fee. I do not understand it all, but I enjoy it.

Maybe their is some hidden risk, but I think being invested in multiple mutual funds across mulitple asset classes will lessen a blow out if it occurs.

 

[ERD50]

And remember this is after paying an additional 2.17% fee on top of the underlying mutual fund fees.

Traditional theory would say they should be underperforming by the fee, and yet they do not, and you can duplicate the strategy yourself without paying the fee. I do not understand it all, but I enjoy it.

Part of the 'traditional' argument is that if a strategy has to overcome that 2% fee, it therefore must be taking more risk. There is some basis for that (especially in fixed income funds), but if the strategy actually *is* providing superior results after fees, maybe that old 'traditional' view is just wrong in this case.

I'm not knocking the approach, it may have merit (which is why I did it for a while). But I am still on the skeptical side that it will provide much boost when you really match it to an index of similar volatility.

BTW, I agree with your earlier idea that we should just measure the negative side of volatility. I'm not aware of any measures like that in the financial world, but they probably exist. I seem to remember looking around for it, but that might have been pre-google days. Seems like just doing the math on the variances that are negative with respect to the target index would do it. You would need a lot of data points.

Maybe someone else is aware of some source of this data?

-ERD50

 

[cute fuzzy bunny]

The newsletter I follow was down by 14.2% in 2002.

Was the newsletter actually in operation in 2000-2002 or is that backtested using their methodology?

 

[Nords]

Geez. I take a couple days off and some of you guys get so far away from the beaten path that we need tracker dogs.

You armchair quarterbacks & backseat drivers: if you read a book and a couple papers before deciding that an investing style is not for you, then that's great. If you conclude from your survey, however, that it doesn't work then you're not doing your research right. Take the positive assertion of the hypothesis, find someone who's made it work, and learn from them. If no such person exists then sure, that style probably doesn't work and your research is probably complete. However I suspect that you will find a person for whom that investing style works-- and you'll conclude that you'd never want to live in their manner.

As for momentum investing, it works at least as well as value investing. Of course it's a lot more effort, but here's some references to look at and draw your own conclusions.

BTW the "No Load FundX newsletter" is one of Hulbert's highest-ranked newsletters:

In addition to his rankings of stock-oriented newsletters, Hulbert also ranks mutual fund letters, with his performance data going back a decade. At the top of the 10-year performance list in total return, and second in risk-adjusted performance, is Janet Brown, editor of No-Load FundX, which uses a momentum-based "upgrading" strategy, in which she moves into no-load funds that have shown the strongest performance over the preceding 1-, 3-, 6-, and 12-month periods.

So while MMND may not be inclined to demonstrate her ability to field-strip a portfolio's asset allocation & correlation while blindfolded, she at least chose to buy the advice of someone who's been able to do it for a while.

Anyway, back to momentum. You could start with Rono at FundAlarm.com who's been a momentum investor almost since he came back from his Marine tour in Vietnam. He's shown me a couple of others who have done well at it. One is Gary Smith, who wrote "How I Trade For A Living". Just reading the first chapter of Gary's recitation of his daily life, where's he's glued to CNBC and hypertrading his IRA mutual funds, will help you decide that the momentum investing routine is not for you. However sparse his life may have been at the time, Gary successfully traded his IRA through 2004. He built it up over a million and ER'd in his mid-50s. Gary is legendary for the clarity of his analysis of technical indicators by demonstrating that they work-- until they don't. Even Gary admits that he can't always tell when an indicator has stopped working. However when an indicator was working he'd frequently put over 90% of his portfolio in that fund and ride it until it trended out. In 2002-4 he made a pile in junk-fund mutual bonds, an achievement which accelerated his retirement.

Rono also recommends Pony Express Bob in this post from Google's cache (FundAlarm.com's board rolls its posts after two weeks). Bob essentially beat Bernstein's ass[-]ets[/-] in the mid-90s, albeit with absolutely breathtaking volatility. Since Bob's board works best in a volatile market, it worked very well for those who had the guts to follow it through 2000-2003. But even I didn't have that level of guts.

Another good board for discussing momentum and technical indicators is FundVision.com. (Gary did a lot of his posting over there but may no longer be reading the board.) The admin, Salil, is a very active momentum trader and has a pretty busy discussion board. However he ruined his credibility by overemphasizing his successes and burying his failures. While I became disillusioned with his integrity, many of that board's members successfully rode his (paid subscription) indicators to beat their asset classes.

For an old-school look at momentum investing read Nicolas Darvas' "How I Made Two Million Dollars in the Stock Market". (This was published over 40 years ago when $2M was actually worth real money.) Darvas was literally dancing his way around the world in the 1950s and trading on momentum trends that he identified from reading stock tickers published in the International Herald Tribune, a method that he admits may not work for everyone.

I tried momentum for a couple stocks in 2001-2. One that worked gangbusters was 4KidsEntertainment (distributor of Pokémon videos & movies) over a four-day period. However much I read and tinkered, I was routinely whipsawed by technicals and sell-stop-losses because I kept insisting on walking away from the computer to [-]go surfing[/-] live my life.

If you're willing to run a daytrader's account with Level II quotes, or if you happen to be able to afford a Bloomberg, then you can probably make momentum investing work for you.

Or, as Bernstein says, you could get a life.

 

[clifp]

Nice summary Nords.

I'll admit that since I don't understand the underlying logic behind momentum investing I'm not anxious to try it. I guess humans are herd animals and we are spooked easily to change course

It appears that all of your examples were for investing in individual stocks. I can understand how that would work for individual stocks. Any studies other than Hulbert's why it works for mutual funds. It seems to me that momentum investing is helped by the financial press [-]porn[/-] saying looked at Google, RIM go!. But there isn't a Cramer out there saying these mutual funds are hot hot hot buy more!

 

[DJRR]

Was the newsletter actually in operation in 2000-2002 or is that backtested using their methodology?

The newsletter was operational in 1999, so those are their actual results before taxes. They also do not include brokerage fees, so the realized return would be lower by however much you have to pay to execute the trades.

They now offer an institutional brokerage account with TD Ameritrade where you can trade all of the recommended funds for $99 per year per account with no minimum holding period.