No. of Recommendations: 21
Greg,

I've greatly enjoyed reading your evaluation. You've broken some new ground, and certainly helped me learn a few things about understanding biotech companies. Now that I've had the time to read your report (well, Part I, anyway) and taken a stab at a couple of new angles, and changed my investments somewhat, I thought I'd post my impressions and even a couple of new twists.

First of all, I've excluded INCY in what follows, for personal reasons (namely, CRA is the better bioinformatics company for my money, and besides, INCY looks like the answer to the question: "which of these biotech companies does not belong?", at least to me). Anyway, INCY isn't in what follows. For the record, I'm long CRA, HGSI, DNA, AMGN, BCHE, IDPH, CHIR, IMNX, MYGN and BGEN. After reading the eval., I sold my MLNM position, and got longer HGSI, DNA, AMGN, BCHE, IDPH, and CHIR, the last three of which I hadn't invested in before.

Second, I do not share your emphasis on Value as an important perspective in determining which of these companies to invest in. If any or all of them cannot deliver on their promise, they will make very poor investments regardless of their current value or market cap. However, they are all possible leaders in (all together now) an important, emerging new industry. Value isn't something we can expect them to exhibit; this is a Growth industry (and I use the term industry very broadly here). So, my guiding principle is more along the lines of: heck with Value, find the company with the best Growth prospects. There's nothing wrong with a value-oriented approach, I just don't think it is, well, of value here. And folks, there's been a lot of talk about larger market cap stocks having less "room" on the "upside." Piffle. Look at CSCO over the last year and tell me mega-caps can't grow. It just ain't so.

Anyway, instead of value, I wanted to get a handle on the effectiveness of the different biotechs in doing their job, i.e., coming up with new inroads to new medicines. The first thing I did was look at their Clinical Index versus their Research Spending. In the table that follows (which does not use the updated CI approach and data, as I didn't have it when I compiled what follows), what I did was assign a 0-5 point value to the CI, and also to their Research Spending/Clinical Index, in an effort to determine which companies were delivering most efficiently. Since I wanted a single number to reflect the companies' strength and efficiency, I then combined the two for a single 0-5 number. The table for this twist on your data is what follows:

RS/CI CI
RS CI RS/CI Points Points RS/CI+CI Research Score
DNA 93 9 10 4 5 9 5
BCHE 21 6 4 5 3 8 4
IDPH 11 3 4 5 2 7 4
GILD 28 4 7 4 2 6 3
BGEN 59 4 14 3 2 5 3
MEDI 10 1 8 4 1 5 3
HGSI 15 2 10 4 1 5 3
MLNM 38 2 17 2 1 3 2
AMGN 198 7 28 0 4 4 2
IMNX 31 2 21 1 1 2 1
CHIR 76 3 28 0 2 2 1
RS/CI Points:	
1 to 5 5
6 to 10 4
11 to 15 3
16 to 20 2
21 to 25 1
26 to 30 0
This score, which is Research Spending divided by Clinical Index, favors companies that bring new compounds to clinical trials for less money; the most efficient companies score highest.

CI Points, and Final Research Score (derived from RS/CI+CI):
9 to 10 5
7 to 8 4
5 to 6 3
3 to 4 2
1 to 2 1
0 to 1 0
This score, for CI, favors companies that bring the most new compounds to clinical trials, regardless of cost.The final column, Research Score above, is the sum of the RS/CI column and the CI column, and therefore gives what I feel is an appropriate, albeit rough, weighting to the various companies regarding their effectiveness in bringing new products to market, certainly an important factor for any of the above biotechs.

*********************

Next I wanted to look at the other major metric for the future success of biotechs: Patents and PCTs. If I follow you correctly Greg, a strong portfolio of patents is important, but the number of PCTs tells us a little more about the future of the company -- and we biotech investors want to be a little more forward-looking. Note in the chart that follows, however, that whether PCTs are given 3 or 5 times their value doesn't matter as far as the rankings of the companies go. Indeed, there is little significant difference how we rank the companies once their Patents and PCTs are combined. Take a look:

Pat PCTs Pat+ Pat+ Pat PCT Pat+ Pat+ Pat+
3PCT 5PCT Points Points PCT 3PCT 5PCT
DNA 553 139 970 1248 5 4 9 5 5
HGSI 112 188 676 1052 1 5 6 5 5
CHIR 327 134 729 997 3 4 7 4 4
AMGN 210 84 462 630 2 2 4 2 2
IMNX 144 48 288 384 1 1 2 1 1
MLNM 51 61 234 356 0 1 1 1 1
BGEN 86 44 218 306 0 1 1 1 1
GILD 63 13 102 128 0 1 1 0 0
IDPH 43 14 85 113 0 1 1 0 0
BCHE 30 11 63 85 0 1 1 0 0
MEDI 3 8 27 43 0 1 1 0 0

Patents PCTs Pat+ Pat+ Points
3PCT 5PCT
200 1000 1300
500-600 167 833 1083 5
400-500 133 667 867 4
300-400 100 500 650 3
200-300 67 333 433 2
100-200 33 167 217 1
0-100 0 0 0 0

In the chart above, all but the first column should be read so that the range is that of the listed cell to the cell above it. That is, 167-200 PCTs=5 Points.The final two columns, which show a score for the combined Patents and some multiplication of PCTs, are identical; I think the winners in this category are pretty clear. For those who prefer a somewhat more narrative approach:

On Patents held, DNA comes out the clear winner, followed distantly by CHIR, and then AMGN. HGSI and IMNX bring up the rear, but the rest aren't even on the list.

On PCTs filed, HGSI is the leader, followed by DNA and CHIR. AMGN follows distantly, the rest very distantly.

If we multiply the value of PCTs and add it to the Patents, we find DNA and HGSI at the top of the list, followed closely by CHIR, then distantly by AMGN, then IMNX, MLNM, and BGEN; the rest don't make the list.

*********************

I don't want to forget the simple metric that I first used to enter biotech investing before I knew a thing about it: Research Spending. Without knowing anything about the companies, I figured that the ones spending the most money on research were the ones most likely to come up with the most, and most important, revenue streams. This is a bit of an Investing in Biotech for Dummies approach, but a year ago it was really the only one I had, and I think it is not without its merits. AMGN is my example: the above two approaches do not flatter the company, yet it is spending more on research than any other biotech. Now, this may be a sign of an inefficient outfit that is about to meet its maker, but I'm betting the other way: I think it is a sign that AMGN may have some new tricks up its sleeve yet. In fact, I went ahead and took a stab at giving the companies 0-5 points as above, and got the following:

Research Spending
AMGN 198 5
DNA 93 2
CHIR 76 2
BGEN 59 1
MLNM 38 1
IMNX 31 0
GILD 28 0
BCHE 21 0
HGSI 15 0
IDPH 11 0
MEDI 10 0

200 Points
167 5
133 4
100 3
67 2
33 1
0 0
Now, about this 0-5 points approach: It must be noted that failure in any one category is not necessarily a foreboding sign that the company is doomed. Far from it, as the approach I've taken here is rough indeed, and quite poor at eliminating companies. Some of the companies that have failed one category here may turn out to be some of the true winners. I would have eliminated any companies that failed all three categories, or even just the first two, Research Characteristics and PCTs and Patents, but there is no overlap. Greg, you picked some good companies to evaluate.

Having said that, however, companies that score highly overall I believe deserve more attention from investors. Take a look at these totals:

Research Score Pat+3PCT Research Spending
DNA 5 DNA 5 AMGN 5
BCHE 4 HGSI 5 DNA 2
IDPH 4 CHIR 4 CHIR 2
GILD 3 AMGN 2 BGEN 1
BGEN 3 IMNX 1 MLNM 1
MEDI 3 MLNM 1 IMNX 0
HGSI 3 BGEN 1 GILD 0
MLNM 2 GILD 0 BCHE 0
AMGN 2 IDPH 0 HGSI 0
IMNX 1 BCHE 0 IDPH 0
CHIR 1 MEDI 0 MEDI 0

All Three: Just Research Characteristics and Pat/PCTs:
DNA 12 DNA 10
AMGN 9 HGSI 8
HGSI 8 CHIR 5
CHIR 7 BCHE 4
BGEN 5 BGEN 4
BCHE 4 AMGN 4
IDPH 4 IDPH 4
MLNM 4 GILD 3
GILD 3 MLNM 3
MEDI 3 MEDI 3
IMNX 1 IMNX 1
Now, the way I see it, the perfect biotech would spend more than any other company, yet it would do so more efficiently than any other company, or at least get the greatest research results for having spent all that money, and in addition, would have the most patents and PCTs. In reality I wouldn't expect such a company to exist, but looking at the above numbers DNA comes frighteningly close to having the best of everything. Among younger, smaller companies, HGSI is the clear leader, and already ties DNA for first place among Patents and PCTs. AMGN is the only company that spends even more on research than DNA, and with DNA is the only company to have a score greater than 1 on each of the three lists. These things are obvious, but for more investing decisions I leave the reader to draw his own conclusions. When I compiled this data I wasn't using the updated material, and with it it looked to me like BCHE had few patents and little money, but was cranking out a very respectable research score. With this company we enter riskier biotech investments. IDPH was like BCHE, only it had come up with less in the way of clinical trials, but with much less money. Efficient, but few paths to success. Risky. CHIR, although well-rounded, didn't have the compounds currently in trials or the patents to compete with the big boys -- or the spending, for that matter. BGEN, well, it was much the same story, and it got worse from there. But those were my thoughts with the old data, and I list them here to give the reader an idea of how I was thinking through these matters. I hope it helps when looking through the new data available now and to come.

Greg, thanks again for giving me something wonderful to chew on, and some badly needed insight that I'm afraid I'm still quite short on. I hope that you find my twists on your data to be of some help, and I'm looking forward to whatever new material you put forth in the future.

Ciao,
Dave
Print the post Back To Top
No. of Recommendations: 0
Simply amazing Dave. If you don't mind, I am going to pick through your post, name some stuff I like and make some suggestions.

...what I did was assign a 0-5 point value to the CI, and also to their Research Spending/Clinical Index, in an effort to determine which companies were delivering most efficiently. Since I wanted a single number to reflect the companies' strength and efficiency, I then combined the two for a single 0-5 number.

I like the five point system thing, it keeps people from splitting hairs. For instance, if one company does 40 million in research and another 39 million, it would be silly to say they are different. Very nice. I think that will go into the next pipeline. Send me your full name so I can add a credit.

The research spending / clinical index ratio is an interesting idea. The problem here may be that the drugs entered the trials on money long spent. Maybe what we can do is determine the RATE of drugs entering clinical trials and divided that by the amount of research spending during that year? I'll keep mulling it over.

If I follow you correctly Greg, a strong portfolio of patents is important, but the number of PCTs tells us a little more about the future of the company -- and we biotech investors want to be a little more forward-looking. Note in the chart that follows, however, that whether PCTs are given 3 or 5 times their value doesn't matter as far as the rankings of the companies go. Indeed, there is little significant difference how we rank the companies once their Patents and PCTs are combined.

I like the backward looking concept of the PCT's. Part of the problem with PCTs that I didn't make clear (for simplicity) in the pipeline is that each PCT may contain several inventions. At least one, but it may have more. I have not figured out how to handle this yet, except to go through each PCT and count the number of inventions, which may not be easy to do. Perhaps I will try to find an average. US Patents, on the other hand, can contain only a single invention. So more PCTs can mean MANY more inventions. Therefore PCTs should probably be weighted like you did.

I would suggest another number, efficiency of research dollars in gaining patents. Perhaps divided the number of patents issued in a year by the amount spend on research dollars that year.

Anyways, I'll keep perusing your numbers. Feel free to drop by again, man. You obviously have a knack for numerical analysis.

thanks,
Greg



Print the post Back To Top
No. of Recommendations: 1
Dave (Wotdabny) wrote:

I don't want to forget the simple metric that I first used to enter biotech investing before I knew a thing about it: Research Spending. 
Without knowing anything about the companies, I figured that the ones spending the most money on research
 were the ones most likely to come up with the most, and most important, revenue streams.


I thought that this would probably just lead to a ranking according to tha size of the company
(which isn't entirely true) and thought to use some ratios instead:
the R&D/revenue ratio and the R&D/MarketCap  ratio.

Why would these ratios mean anything?

Looking at plain R&D and scoring for its size assumes, as you state.
that every research $ buys the same amount of "revenue stream"
However, a given rev. stream will be of more or less importance for a compnay depending on its size. 
Therefore the RD/Revenue ratio might give us a hint how quickly the company
might grow its rev. compared to the others 
(under the assumption that each research $ buys the same amount of revenues, which is of course questionable).
My biggest objection to the use of this quantity would be that revenues are no meaningful number for companies
 such as HGSI, that haven't even started making money the way they eventually intend to - ideally, they'd
 have zero revenues and therefore infinite score.
Probably one should compare only companies at a similar stage that way. The second objection would be, 
that a large company as AMGN could probably generate more sales out of the same discovery as BCHE could, thanks to superior brand and marketing.

The RD/MCap ratio tells us how expensive those research $$$ that we buy are and thus how promising the ongoing research is considered to be.


Do you think these ratios have anything meaningful to say?


Here are the 11 companies, sorted by R&D expenses (FY1999 data) with the other data besides:

Ticker	R&D(M)	Rev(M)	Mcap(B) RD/Re  RD/MC P(RD) P(RD/R) P(RD/M)  	
IDPH	 42.8	 118	 5.6	0.36	 7.64	0	1	2	
BCHE	 59.3	 196	 3	0.3	19.77	0	1	3	
MEDI	 59.6	 338.4	11.5	0.18	 5.18	0	0	1	
HGSI	 60.6  	  24.5	 9.3	2.47	 6.52	0	5	1	
GILD	112.9	 169	 3.4	0.67	33.21	1	2	5	
IMNX	126.7	 243.5	31.8	0.52	 3.98	1	2	0	
MLNM	159.9	 183.7	10.1	0.87	15.83	1	2	3	
BGEN	221.2	 794.4	16.4	0.28	13.49	2	1	3	
CHIR	303.4	 421.7	 8.2	0.72	37	3	2	5	
AMGN	822.8	3340.1	72.8	0.25	11.3	5	1	2

and here the orderings accordingto the three numbers (lowest being most attractive):

R&D	RD/Rev	RD/Mcap
IDPH	MEDI	CHIR
BCHE	AMGN	GILD
MEDI	DNA	BCHE
HGSI	BGEN	MLNM
GILD	BCHE	BGEN
IMNX	IDPH	AMGN
MLNM	IMNX	DNA
BGEN	GILD	IDPH
CHIR	CHIR	HGSI
DNA	MLNM	MEDI
AMGN	HGSI	IMNX
Print the post Back To Top
No. of Recommendations: 0
Excellent observation. In the 'Pipeline report I actually use the inverse of R&D/Mcap and called it Price to Research ratio.

In this case:

P/R = market cap / R&D expenditures

You can conceptualize it similar to P/E, where you are paying X dollars in market cap for Y dollars in R&D.

It is the Price you are paying for each dollar of R&D. Think of every research dollar as fungible, i.e., not materially different from one company to the next. Then, you want to buy as much research for your stock purchase as you can.

I'll throw these numbers up here when I get the chance.

Greg
Print the post Back To Top
No. of Recommendations: 0
Thank you for your post:
Subject: New Twists on Your Data
Author: Wotdabny Date: 2/15/00 4:08 AM Number: 72

It was very informative. I'm trying to do a historical study of R&D budgets in the biotech industry, and was wondering where you got your R&D budget data?

Thanks.
Print the post Back To Top
No. of Recommendations: 0
Thank you for your post:
Subject: New Twists on Your Data
Author: Wotdabny Date: 2/15/00 4:08 AM Number: 72

It was very informative. I'm trying to do a historical study of R&D budgets in the biotech industry, and was wondering where you got your R&D budget data?

Thanks.
Print the post Back To Top
No. of Recommendations: 0
Actually, he got the data from the "Harvesting the Human Genome" report on evaluating biotech pipelines (hit the link below this message).

The data that is in the report is gleaned from annual and quarterly filings of each company. Slow and tedious, but it got the job done.

I think part of the difficulty in doing what you want to do is the short lifespan of many of these biotech companies. The research expenditures tend to not be realized as actual gains until 10 to 15 years later.

I also fear that the predictability of finding blockbuster drugs is so low that it will mask the significance of research expenditures.

I am curious as to what you plan to correlate these expenditures to - stock price, number of successful drug candidates, earnings, etc.

Again, I can work with you on the backtesting if we can isolate perhaps the 5 or 10 largest biotech companies. Also, we need to define very clearly what we are looking for.

thanks,
Greg

Print the post Back To Top
No. of Recommendations: 0
My earlier question about where to get research spending figures was a dumb one, as it turns out-- I wasn't looking at the right place in the 10k's. A co-worker who is a finance major set me straight on that. Oops.

Anyway, what I'm looking for are the best predictors of (%change_stock_price - %change_s&p500) for the following year.
Print the post Back To Top