Qcom analysis by Rich the Janitor, via Ruffian (Ruffian, Thanks in advance)....
Pieced together a 3-part post from one of the Q threads; (stormy seas over there):
"I will try and be very conservative. So the story goes:
How to do you place a valuation on a company that is predicted to grow it's yearly EPS by 50% a year for the next five years? Even harder is to try and valuate the same company when it continues to blow those 50% projections out of the water. As much as I would love to speculate about HDR, China and WCDMA, the truth of the matter is that all that 'is' speculation. Although I buy speculative stocks, I would not invest 30% of my portfolio in them. When I first bought Q it was speculative and 1% of my portfolio. As I figured out its' earning path, it went from speculative to medium risk in a short time. As did the increase in my position. So, for today, let's focus on what is real. what is in the cards for Q over the next three years? I am using three years because I don't see any 3G alternatives before then. Thus, we can draw concrete conclusions over that period of time. So here are the factors that contribute to Q's value.
1. P/E
I won't do it here, but I can debate with anyone that a company that exceeds EPS growth of 50% a year for the next three years deserves a P/E of 100+. IF you disagree, let's debate P/Es in a separate post. IMO, as long as Q grows EPS by 12-20% per quarter, there is no P/E that is too high.
2. CDMA growth
Typically, I would include earnings after P/E, but because Q's earnings have mirrored (actually exceeded) CDMA subscriber growth, I feel it is logical to include this one before we discuss earnings. In December of 1998, all the major communications analysts called for CDMA to grow by 35-50% for the next 5 years. Well, as we know. They were a bit conservative. They underestimated two things. The first and most important was the growth of wireless. The second was
QCOM's ability to sell and grow worldwide CDMA acceptance. In December of 98, there were a total of 23M worldwide CDMA subs - an increase from 8M the year before. In Q3 (jul-sept) of this year, 7.5M subs were added. A staggering
increase of 22% over the previous quarter. This put total subscribers at 42M. Almost double from the year before with still our best quarter in fromnt of us. The pros had called for about 35M by end of 1999 and it look as if we will finish closer to 48-50M. Expect a slowdown? Not at all. A 50% increase in the year 2000 would represent an end of the year number of 75M subs or 25M new subs in 2000. Need 112M in 2001 and 160M at end of the year 2002. 2001 represents the commercial launch of IXRTT (2.5G) in Korea and North America which increase CDMA equipment sales significantly, although I will not speculate on how this will increase sub growth. Not because I don't think it will (IT WILL), but because I would like to continue focusing on the conservative. So, keep track of the CDMA growth numbers. They have always been the key. Why? Two reasons:
?® Because QCOM is paid an ASIC royally and a phone royalty on every CDMA phone sold, their EPS growth can conservatively be drawn in parallel
?® The biggy. CDMA subscriber growth is the key, but CDMA replacement phones are the reason that you have a gold mine in your portfolio. As the total number of subscriber 'base' grows, the number of replacement phones tells a new story. QCOM owns the CDMA ASIC world. In a WORST case scenario, QCOM may have only 50% of the CDMA ASIC market. Although I think that 70-85% is a more realistic worst case scenario, I want to be very clear. 50% of the market will allow QCOM to continue EPS growth 'well' above 50% a year. But we don't have to worry about losing ASIC share. Why? Simple - QCOM just announced successful trials of their 1XRTT chip and software. To date, MOT, Samsung, DSPC, LSI or Nokia have not been able to develop a chip that is half as competitive as the MSM. In fact, half of those companies don't even have chips in high volume manufacturing. With the exception of MOT and NOK who have jokes for ASICs, everyone else is stillwaiting for HV. So as they finish up their inferior chips that have higher cost and analog-like performance, QCOM is sampling a product (1XRTT) that represents a two year development headstart on a CDMAOne chip.
3. Earnings
It is all in the earnings! Now that we have a model, a pattern if you will, of CDMA growth, we can determine QCOM's part of that growth and how it relates to their growth. An equation for this could be:
E= S + R
E= Yearly EPS Growth
S= subscriber growth in percentage
R= replacement phones (30% of total subscribers)
Thus in 1999,
E=1.0 + .3
E=130%
And in 2000 a conservative view is:
E= .5+ .3
E= 80%
Using this to determine actual growth variables, one could use the equation,
T = A + (E)A
T = total phones sold in the year
A = new subscribers in the year and
E = EPS growth
You can see how the base subscriber rate acts as a huge growth multiplier for replacement phones. You math majors could perhaps think of a better equation with actual variables, but this is the best I can do with my janitorial-like mind and half a bottle of Chianti. But understand this: The more CDMA subscribers there are, the more replacement phones that will be sold. Easy enough so let's finish by doing the numbers and let's be as conservative as humanly possible. A worst case scenario.
Q1 = .28
Q2 = .50 (I added a premium because HS division is gone)
Q3 = .56 (12.5% growth)
Q4 = .63 (12.5% growth)
2000 earnings = $1.97
so using a 100 forward PE, TODAY we should be at $197/share in a VERY conservative way. If $3 ($12 pre) is 2001 earnings, we should be at $300 a share in December of this year.
Remember these facts:
A) I am only using subscriber growth in parallel to EPS growth. I DID NOT use the more accurate equation above
B) I am not using a premium for HDR, China, WCDMA or Huge expected data growth
So what if we use Janitor's numbers:
Q1 = .28
Q2 = .50
Q3 = .60 (20%)
Q4 = .72 (20%)
Or earnings of 2.10 so $210/share today and 357/shr in December PLUS I would add FMV or $25/share for HDR, $25 per share FMV for China and $10/Share for WCDMA for a total of $417 at year end. I used less WCDMA because it is the
furthest out there in terms of definitive analysis.
Am I still being conservative? YES! What about other CDMA applications and other Q business units? $25 premiums are cheap for the like of what China would do to our growth rates. What about HDR and new licensing fees for IS-2000? What about new data users? Well, as I said at the beginning: Janitor doesn't want to speculate. I want a sure thing and right now QCOM is a sure double and likely triple based on CDMAOne growth and earnings alone.
Observations.
1) I agree that replacement phones estimates are low, however I would like to err on the conservative side. In the US, phones are replaced every two years where in Korea, it is average to buy a new phone 2-4 times a year for the latest and greatest model.
2) Regarding the risk of Q losing market share. At the end of the analysis I used conservative numbers that assumed Q would lose market share, although I don't think they will. Their recent sell to KYO increases their market share, in fact. If Nokia starts buying chips............watch out. Anyway, Q currently has > 90% of CDMA ASIC market. My conservative assumption made this number 50% and my 'Janitor' assumption used 70%. So my model did clearly take this into consideration.
3) ASP - good one. I will debate against ASP erosion, although the truth is that I forgot it in my model. :-). UNTIL a competitor arises, the ASP on a 2g (IS-95A/B) ASIC will not go down. Maybe a small percentage to keep customers smiling. 20%, no. When a proven alternative solution is in HV production, I believe the ASP could decrease 30%, but (BIG BUT) Two things to remember:
?® QCOM is introducing new chips that are more highly integrated, backward compatible, and higher performance. They are on their 7-8 generation MSM and the competitors are coming out with their first. So ACME will come out with a IS-95A which current consumption allows for 90 hour standby time and 2 hour talk-time, Q will have an MSM that provides GSM like numbers, IMO
?® Rule number 1 in this business is to not piss off your supplier. Anyone who wants a IS-2000 chip on time and for a reasonable price will be using IS-95A/B MSMs, imo. If I'm Samsung, I don't want LG getting first priority on chips at a lower cost. The fact is that consumer demand for lower cost phones with higher battery life and more features will keep Q's ASP from decreasing at 20%. Although I agree with you that a 20% erosion should be used in the equation.
4) last item - My premium on Q2 earnings. I really don't know how to put a definitive number on Q1 and Q2 earnings. I have no idea what impact the subscriber division had on earnings for this quarter and I have no idea what impact the sale of the unit will have on next quarter. My hunch is minimal impact on Q1 earnings and Huge impact on Q2. Interested to see what the Pro Forma numbers are. "
- Janitor |
