Magic Formula - Joel Greenblatt

**blast_investor** Wed Dec 07, 2005 10:25 pm

Magic Formula - Joel Greenblatt
Joel Greenblatt is famous value investor, hedge fund manager of Gotham Capital. His past 20 years of performance of his hedge fund was 40% annualized return. He is so rich that now he returned all outside external money and only run his own money for the fund.

Greenblatt's main method is long term holding of focused portfolio of 5 to 10 stocks, strictly with value method: low valuation with high earning yield, and high return on asset.

Here is his audio talk on his method:

http://www.npr.org/templates/story/story.php?storyId=5036846

**blast_investor** Mon Jan 09, 2006 2:18 am

Greenblatt's Personal Investment Process

Looks for value with a catalyst, so nice things happen sooner. Special situations are just value investing with a catalyst. They are simply different places to find cheap stocks.

In his own hedge fund, Greenblatt uses the basic principals in the magic formula: Look for high ROC and high earnings yield.
Tries to figure out what "normalized earnings" will be 3-4 years into the future.

Makes sure the stock is very cheap based on normalized earnings.

5 to 8 securities can make up 80% of his portfolio. One position could be up to 30%.

Having a concentrated portfolio works well for lazy people. Not that many stocks to track.

Thinks about how much he could lose if he's disastrously wrong.

No formal process or time frame for purchase decisions. Usually spend one month or so to do research. In difficult situations for which he and his partners have time, research could take months.

If there is a great opportunity which, in their opinion, won't last, and if they feel they understand it, they sometimes use the approach of "Ready, Fire, Aim!"

Has financials and utilities in the portfolio.

EBITDA Minus Maintenance CapEx
For his own investment practice, Greenblatt uses a different input for earnings.

He thinks that "EBITDA - Maintenance Capital Expenditure" would be a better measure of earnings power, but it can be difficult to calculate.

On Long-Short Strategies

Q: "How about long the top deciles of cheap stocks and short the bottom deciles of expensive stocks?"

A: "I am not a fan of shorting. The long-short guys blow up every eight years. I call it the 'I got it! I got it! I ain't got it!' Strategy."

Fair Bet Yet Unfair Investing

Look for a big mess that seems too complicated, not well understood, not well followed, and requires too much work. People don't want to do the work, but once you do the research, you will be at an advantage.

Look for semi-complicated situations. The key is to identify what cuts to the core.

The Macro Picture Is a Distraction

In 1999 and 2000, there were plenty of non-internet values out there. If you were worried that the burst of the internet bubble would have dragged those values further down, you would have made a big mistake. "Fortunately, we ignore the macro picture," said Greenblatt.

In 2002 and 2003, there were plenty opportunities in small caps. If you were concerned that the bear market could go on further, you would make another mistake.
"Now the value is in big caps. But if you look at the macro picture, the consumers could drop dead, the housing bubble could drag everybody down?- We ignore those. We have no macro view," said Greenblatt.

Everything is cyclical. Values can always be found somewhere.

Ignore Volatility and Stock Prices

Taking your clue from the stock prices is crazy. If you could value companies, you should ignore the noises from volatility and stock prices.

Things such as volatility have nothing to do with buying a good, cheap company for the long run. Greenblatt said, "It is kind of ironic that, the older I get, the longer time horizon I look at." (Laugh)

The Efficient Market Theory is a crazy way to look at the market. "Pick and choose. You can beat the market. It is worth the work."

The Vogue of Return on Capital

There seems to be a movement towards high return on capital.

The low P/B stocks haven't work very well in the past 10 years.

Don't know if or when this trend will reverse.

--------------------------
Dr Brian Zen has wonderful note on Mr Greenblatt talks, to read full article:

http://www.gurufocus.com/news.php?id=802

**springsnailt** Thu Mar 09, 2006 2:16 pm

Greenblatt's magic formula seems to be oriented for non-enterprising investors.

Blast, do you pay attention to picks from the formula? Here's a list of 25 with MktCap>100M

Name (in alphabetical order) Ticker Market Cap
($ Millions) Pre Tax
Earnings Yield Pre Tax
Return on Capital
Amern Eagle Outfitters Inc AEOS 4,209.19 13% > 100%
Freightcar America Inc RAIL 868.16 10% > 100%
Ftd Group Inc FTD 269.01 10% > 100%
Grubb & Ellis Co GBEL 110.02 20% > 100%
Housevalues.Com-Redh SOLD 228.19 13% > 100%
I2 Technologies Inc ITWO 346.55 14% > 100%
Intersections Inc INTX 152.30 19% 75 - 100%
Intervideo Inc IVII 147.78 22% 75 - 100%
Jakks Pacific Inc JAKK 659.78 17% > 100%
K-Swiss Inc -Cl A KSWS 984.68 14% > 100%
King Pharmaceuticals Inc KG 4,250.90 15% > 100%
Korn/Ferry International KFY 842.41 10% > 100%
Kos Pharmaceuticals Inc KOSP 2,043.83 11% > 100%
Lennox International Inc LII 2,266.05 20% 75 - 100%
Magellan Health Services Inc MGLN 1,357.94 14% > 100%
Mannatech Inc MTEX 354.46 12% > 100%
Marvel Entertainment Inc MVL 1,659.79 10% > 100%
New Frontier Media Inc NOOF 166.04 11% > 100%
Pinnacle Airlines Corp PNCL 160.64 36% > 100%
Portalplayer Inc PLAY 555.75 12% > 100%
Pw Eagle Inc PWEI 237.35 33% 75 - 100%
Vaalco Energy Inc EGY 401.96 16% > 100%
Vertrue Inc VTRU 429.86 10% > 100%
Westwood One Inc WON 952.28 11% > 100%
World Air Holdings Inc WLDA 203.67 30% > 100%

**noguessinginvesting** Sun Mar 12, 2006 9:22 am

Magic Formula Process
So, I read the book, and the step-by-step process from which Mr. Greenblatt suggests buying the magic formula stocks, but had a question. After selecting the market capitalization, and selecting the top 25 ranked stocks according to his formula .... how do I select the first few to invest in? Should this be a random selection of the top 25? Looking at some of the graphs of the underlying securities - I found that many of them were overextended - when I would typically wait for a pullback to purchase any shares. Would I be better off selecting the equities that are currently at or below their 100 day MA for example? Any advice would be greatly appreciated. Thanks! Noguessinginvesting

**blast_investor** Sun Mar 12, 2006 12:16 pm

Re: Magic Formula Process
Hi noguessinginvesting,

I have read Joel books as well. I am not trying to recommend his approach for real stock investment.

But, if you are trying to pick a few from his 25 stock list, then you are not following Magic Formula. Joel's magic formula does not know how to pick a few from 25.

noguessinginvesting wrote: So, I read the book, and the step-by-step process from which Mr. Greenblatt suggests buying the magic formula stocks, but had a question. After selecting the market capitalization, and selecting the top 25 ranked stocks according to his formula .... how do I select the first few to invest in? Should this be a random selection of the top 25? Looking at some of the graphs of the underlying securities - I found that many of them were overextended - when I would typically wait for a pullback to purchase any shares. Would I be better off selecting the equities that are currently at or below their 100 day MA for example? Any advice would be greatly appreciated. Thanks! Noguessinginvesting

**blast_investor** Sun Mar 12, 2006 12:18 pm

Hi springsnailt,

My approach is more of pure Graham style: digging for deep value.

I studied Magic Formula 25 stock picks. They do not meet my criteria.

springsnailt wrote: Greenblatt's magic formula seems to be oriented for non-enterprising investors.

Blast, do you pay attention to picks from the formula? Here's a list of 25 with MktCap>100M

**noguessinginvesting** Sun Mar 12, 2006 12:57 pm

Hi Blast,
Thanks for the reply. I wasn't clear in my question. I intend on purchasing 4-5 stocks (out of the 25 top choices) every quarter as recommended in his book. So over the course of a year, I'll have the full 20-25 stocks that he suggests. However, I'm wondering if I'd be better off selecting stocks that technically have not had recent price advances - trying to eliminate purchasing an overbought issue? What do you think? Noguessinginvesting

**blast_investor** Sun Mar 12, 2006 5:31 pm

I do not know this.

Over the long run, how you do that would not matter that much.

noguessinginvesting wrote: Hi Blast,
Thanks for the reply. I wasn't clear in my question. I intend on purchasing 4-5 stocks (out of the 25 top choices) every quarter as recommended in his book. So over the course of a year, I'll have the full 20-25 stocks that he suggests. However, I'm wondering if I'd be better off selecting stocks that technically have not had recent price advances - trying to eliminate purchasing an overbought issue? What do you think? Noguessinginvesting

**nodoodahs** Thu Apr 06, 2006 12:45 pm

noguessinginvesting wrote: Hi Blast,
Thanks for the reply. I wasn't clear in my question. I intend on purchasing 4-5 stocks (out of the 25 top choices) every quarter as recommended in his book. So over the course of a year, I'll have the full 20-25 stocks that he suggests. However, I'm wondering if I'd be better off selecting stocks that technically have not had recent price advances - trying to eliminate purchasing an overbought issue? What do you think? Noguessinginvesting My personal opinion is YES. Limit the value buying from "magic formula" to stocks trading under their 200 DMA, or those with 52-week returns of 0% or less. Magic formula will tend to catch cyclicals at their peak, you don't want to be buying petro shipping companies in March of 2005, but they were all over the Magic Formula at that time.

**blast_investor** Thu Apr 06, 2006 1:08 pm

Once you apply your TA criteria to Magic Formula picks, it is no longer the "Magic Formula" that Joel Greenblatt recommends. This is new hybrid method.

Joel Greenblatt used more than 20 years of back testing data mining to prove that his "magic formula" works. However, nobody back tested this new hybrid method.

nodoodahs wrote: My personal opinion is YES. Limit the value buying from "magic formula" to stocks trading under their 200 DMA, or those with 52-week returns of 0% or less. Magic formula will tend to catch cyclicals at their peak, you don't want to be buying petro shipping companies in March of 2005, but they were all over the Magic Formula at that time.

**nodoodahs** Thu Apr 06, 2006 4:13 pm

"This is new hybrid method" - well, yes. noguessinginvesting asked about that type of hybrid method, and I answered. I assumed since he posted it on an open thread he might expect an answer from any registered member of the forum, so there it is.

I backtested on a 5-year time period for a combination of PE, PB, DE, and 52-week return, using data I pulled from getting a temporary "Value Line" subscription. The subject group had higher average return, higher pecentage of stocks with positive return, and higher percentage of stocks with 25% or higher return, given a one-year holding period.

With a little time, some money for the data purchase, and some skill with MS Access or other database product, you can do the backtesting yourself. It ain't rocket science.

It also makes intuitive sense that peaking cyclicals would have high ROIC and Earnings Yields for the most recent year (which is what you get from the Magic Formula site). If you want to avoid buying a peaking cyclical, one way (not the only way, but one way) to do that is to use 52-week return on trading below 200 DMA.

In practice, JG and his fund will smooth and normalize earnings, so the ROIC and EY they use is not what's on the sight or what you or I could find through a typical screener. They also review proxies, K's and Q's to make sure the companies are clean, and probably do other earnings quality metrics checks. These are all steps one can use to narrow down the list.

If you use the screener list, the only options you have are 25 (smallest grouping) and market cap. To limit yourself to 4-5 per quarter, you need some methodology that hopefully isn't arbitrary. My TA suggestion, while not published in a book or research paper and not backtested over 20 years, has been backtested over 5 years and makes intuitive sense. Further, review of the filing documents is another way to narrow down the list to five. Finally, you could use a group of financial metrics based on solid research to eliminate companies with issues. If you want some information on the financial metrics that I've researched, you can find them here.

http://nodoodahs.com/investing-strategies/metric-system/

**blast_investor** Thu Apr 06, 2006 5:53 pm

Well, that is different issue. noguessinginvesting question was really about gradually building up the full magic formula portfolio. Therefore, the topic was still magic formula.

Eventually the portfolio will be 25, it really does not matter that much for the initial built-up (say 1 year). After all, after a year, the portfolio will be just like 25 stock picks full Magic Formula portfolio. If one use chart for initially build-up, it does not matter that much because after 1 year, the portfolio should be 25 stock picks of magic formula picks if that is what investors wants. The chart only works for the initial short time frame.

"Magic formula" is very blind method, it does not know which out of 25 are good, which are bad. 25 stock in a portfolio is pretty large number for diversification purpose to insure the portfolio to work. There might be some picks in magic formula picks to go down to zero, it is all possible, but the portfolio will be fine according to Joel Greenblatt's research.

I assume the hybrid method you mentioned will not invest into full 25 list of MG picks eventually. Once an investor pick and choose from magic formula picks using the hybrid approach, it is no longer the "magic formula" portfolio. Furthermore, if the stock picks number drops down significantly below 25, say 10, the portfolio would be very focused and no longer the same "diversified" portfolio as MG portfolio is. Now we are talking something that has nothing to do with Joel Greenblatt. The worst nightmare would be, the investor would pick and choose the bad picks in 25 stock picks that will go down dramatically and avoid the magic formula stocks that would rally big down the road.

Of course, if investor does not want "magic formula", and only wants the "hybrid approach", that is different issue.

nodoodahs wrote: "This is new hybrid method" - well, yes. noguessinginvesting asked about that type of hybrid method, and I answered. I assumed since he posted it on an open thread he might expect an answer from any registered member of the forum, so there it is.

**nodoodahs** Thu Apr 06, 2006 6:58 pm

blast_investor wrote: noguessinginvesting question was really about gradually building up the full magic formula portfolio. Therefore, the topic was still magic formula. Actually, blast, noguessinginvesting's question was specifically regarding TA as a method of further screening the magic formula picks. noguessinginvesting wrote: Would I be better off selecting the equities that are currently at or below their 100 day MA for example? noguessinginvesting wrote: However, I'm wondering if I'd be better off selecting stocks that technically have not had recent price advances - trying to eliminate purchasing an overbought issue? That being straightened out ...

Assuming you are using the Magic Formula site to initially screen your purchases, the site limits to 25 as the smallest output. Therefore, unless you intend to purchase 25 stocks at a time, you need to find a way to narrow the list down. There are several approaches one could take, including:

1. a review of SEC filings (proxies, K's, Q's) which is something that JG does,

2. a review of financial metrics which is also something that JG does, although they may not do it exactly as I outline in my post http://nodoodahs.com/investing-strategies/metric-system/

3. random methods, and

4. technical methods.

In my opinion, and in my own backtesting, I have found that adding a component of 200 DMA or 52-week gain improves the results.

Blast, you're a data-miner. Why be so hostile to the idea? Why not try adding it to your screening process and see how it works? You could backtest it on your own database.

**blast_investor** Thu Apr 06, 2006 7:31 pm

Well, It is not that I am "hostile". I do not believe a back testing of 5 years means any thing. The data mining effort of back testing 20 years to 50 years would be a full time job, a job that is probably most suitable for an economic professor in University. But only back testing 20 years to 50 years would yield any somewhat "proven" method that can make some money.

Personally, I mainly piggyback the established proven methods. I am not looking for or inventing new method. Most of well-known value investors are piggy-backing Ben Graham anyway.

In my opinion, making money with an established value investing approach is easier job. It is very hard to find any new approach of value investing. Trying new approach is particularly not suitable for average inexperienced investors.

nodoodahs wrote:
Blast, you're a data-miner. Why be so hostile to the idea? Why not try adding it to your screening process and see how it works? You could backtest it on your own database.

**noguessinginvesting** Thu Apr 06, 2006 9:00 pm

thanks for the replies
I appreciate your responses. Nodoodahs ... what did you mean by the 52 week gain screen? Is that a way of measuring where a stock's price is in relation to its 52 week history? A recent Barron's questioned the methods of the magic formula's backtesting. However, although they weren't able to reproduce the same results - the results they achieved still beat the overall market.

As far as application of your advice ... I wouldn't mind using a screen to pick the best short term candidates when I purchased 4-5 stocks every few months, until finally reaching the 25-30 that they suggest. I would probably have to run a top 50 screen on their site in order to get more stocks that are diversified in their sectors as well (so that I wouldn't buy a bunch of oil companies, etc). Thanks, noguessinginvesting