Description

Intuitive Surgical (ISRG) makes a great robotic surgery product that very likely provides a better patient outcome in prostatectomy, cervical and/or uterine cancer, sacrocolpopexy, myomectomy, complex hysterectomy, as well as certain cardiothoracic and general surgeries. That being said I think a short position in ISRG has an upside return of ~45% (if just part 1 of my thesis plays out) and possibly much more if the 2nd part of my thesis is correct. The cost to borrow is zero, only 6.2% of the float is short (as of 3/15/10) and it trades over $180M per day (based on the daily share volume over the last 3 months).

The way I view ISRG's business is that it has 2 segments that each make up about half of the company's revenues and profits. One is Systems, which is a flat to declining business (grew just 0.9% in 2009) and the other is Instruments & Accessories/Services & Training, which is growing at a high rate (50.7% in 2009). However, combined the top-line and operating profit growth YoY in 2009 was in the 20-21% range, declining from 50-60% in 2008. I believe that revenues and profits will grow, but at a continued decelerated rate of 15-20% in 2010, then falling to low double digits thereafter (2011-2013) and then having mid-to-high single digit EPS growth thereafter. At ~45x 2010E EPS (backing out net cash), a hyper growth rate must last a much longer time to justify the valuation.

There are a few main points to the short thesis (in order of importance):

1) I believe systems revenue has a major long-term headwind that will cause unit sales to be flat to declining for a number of years. The key metric to focus on is product utilization--defined as the avg. # of procedures per machine/total # of possible procedures per machine. In Q4'09 there were 58,000 total procedures performed on an avg. installed base of 1,351.5 machines--or 42.9 procedures per machine over 13 weeks = 3.3 procedures per machine per week = 2/3 of a procedure per day (assuming a 5-day work week).

As for maximum utilization per machine I've come at this 2 different ways and both get me to approximately the same answer, which is 10-12 procedures per machine per week--or about 2 per day (but 3 per day is done at several high volume hospitals). First, JP Morgan did a survey (6/24/09 report) of urological and gynecological surgeons that use the machine and found that the median response to the number of possible procedures per machine per week was 12. In fact, urologists on avg. believed it was 9 per week (due to longer procedure times) while gynecologists on avg. thought it was 15. This is important because over 75% of all prostatecomies are already done using a da Vinci so a majority of the growth must come from gynecological surgery, where the capacity per machine is believed to be 15 per week vs. 3.3 procedures currently performed.

I recently conducted a proprietary survery of 52 (35+ years old) da-Vinci trained gynecological surgeons from hospitals (with 200+ beds where 90% of its machines have been placed). First, I asked these surgeons to define their usage. 17% said high (8+ procedures per week), 62% said medium (4-8 procedures per week) and 21% said low (<4 procedures per week). Second, I asked what is the minimum threshold of surgeries per machine per week needed to prompt his/her hospital to buy another da Vinci robot. 65% responded 8+ procedures per week, 10% said 4-8 per week and 25% were unsure. However, there are some other contributing factors as to why additional robots are purchased despite under utilization. 35% said it is so that departments (i.e. urology and gynecology) don't have to schedule around each other's surgeries, 50% said due to charitable donation, 50% said to attract talented surgeons, 6% said if departments are on different floors, 6% said if it's building an additional hospital, 6% said used for teaching residents and 15% said that there are no other reasons besides utilization that would cause his/her hospital to buy another robot.

I believe in tough economic times (combined with potential negatives from ObamaCare) hospitals must really question if they can get their ROIC up by simply doing a better job of scheduling (to get utilization per machine up dramatically) as opposed to just buying a second machine because two departments don't want to deal with the extra difficulties of working together to efficiently schedule operating room time. No matter how you look at it the, current avg. utilization of these machines is at best just 33%. A gynecological surgeon at a major hospital in NYC said that his hospital required the urology and gynecology departments to work together to get utilization per machine up to 12 procedures per week before they could ask the hospital to get a second machine. That hospital ultimately achieved that goal by taking utilization up to 11-12 procedures per week.

The bulls believe that the reason systems sales were flat in 2009 was entirely due to the recession. While I don't disagree that the recession had some impact in 2009, I believe that this economy will force hospitals (as it has with consumers) to get more bang for their bucks and stope being so wasteful.

Bulls also believe that a replacement cycle will happen, which will boost system sales. While that certainly could happen at some point, I believe we are many years away from it occuring in any material way. If the usage on these machines continues to be relatively light I don't see it happening. The hospitals believe their machines will last at least 7 years. Keep in mind that the avg. age of its machines in the field are just 2.8 years old. At the same time, if a hospital does upgrade it will sell its machine back to the company who will then sell it to a Tier 2 or 3 hospital or sell it into the international market.

 
2) Company guidance is for approximately 95k hysterectomy procedures to be performed on a da Vinci in 2010--up from 69k in 2009. In its Q4'09 presentation it stated its realistic long-term U.S. goal was 450k gynecological procedures vs. 350k in Q4'08--an increase of ~30% in just 1 year. While its urology target of 150k procedures remained the same vs. 1 year ago, it lowered its cardiothoracic goal by 20k down to 100k procedures, but added a general/H&N category of 200k that didn't exist one year ago. Also, it increased its international goal from 620k to 900k procedures. All in all it increased its target market from 1.24M annual procedures to 1.8M--an increase of 45% in just 1 year! This compares to its guidance of 277k procedures in 2010 and 205k performed in 2009--an increase of 45% in just 1 year!

I find it hard to believe that this is realistic. Call me skeptical, but I think it has something to do with the fact that a year ago the stock was at ~$100 and management was given stock-based compensation of $97M in 2009. As it continues to dole out an increasing amount of stock-based comp., I think management wants the market to continue to afford ISRG the same or higher P/E multiple on future EPS.

Gynecological Target Market is Likely Materially Less than ISRG's Goal:

I asked the 52 surveyed da-Vinci trained gynecological surgeons how many robotic procedures they currently perform per month and what their estimate was for 2010, 2011 and then long-term (2012 and beyond). The weighted avg. response was 3.48 procedures per month currently, 4.73 for 2010E, 5.50 for 2011E and ultimately reaching 6.56 over the long-run. That's an increase of 88.4% over the long-term--a far cry from the 650% increase the company would have you believe.

My individual channel checks (chiefs of surgery at major hospitals and gynecological surgeons) believe the total U.S. potential for gyencological surgeries is more like 185k. The reason is that only certain types of gynecological procedures truly benefit the patient using a da Vinci and it costs the hospital more money to do the surgery using the da Vinci (with no increase in payment/reimbursement). The ultimate # of procedures that will be done using a da Vinci is very hard to prove and is a distant #2 to the thesis. Over the long-run I believe this will play out, but I don't expect this to come through over the next year given that the procedure growth is still going to be ~35% per company guidance and surgeons surveyed.

 
3) The avg. revenue per gynecological procedure is less than that of a urological one. So while a certain amount of unit volume growth is likely to come through over the next few years, the total revenue growth is unlikely to be as good. One can see this from the YoY decline in revenue per procedure (instruments & accessories revenue/# of procedures) in 2008 and again in 2009--down -4.5% and -11.8%, respectively. The % mix of procedures in 2007, 2008 and 2009 made up by gynecological was 16.0%, 25.0% and 33.7%, respectively.

Sum of the Parts (conservatively allocating a majority of R&D to no/low-growth Systems):
25x 2010E Instruments & Accessories/Services & Training Op. EPS ($4.54): $113.51
15x Q4'09 Run-Rate Systems Op. EPS ($3.28): $49.15
Cash & equivalents (as of 12/31/09): $30.75
Total ISRG Value: $193.41 (45% downside)

 
Based on my 2nd and 3rd points, there is a risk to growth slowing in the Instruments & Accessories/Training & Services segment, which calls into question whether a 25x 2010E P/E multiple is even justified. Thus if my 2nd point plays out then there is much more downside than my sum-of-the-parts, but since it is more difficult to prove I don't count on that--rather view it as an extra put option.

Risk (timing): I have no idea when ISRG might miss earnings. Given that procedure growth is likely to be good in 2010, it could be enough to give ISRG the growth the market is looking for and thus the stock price could very well increase more this year. That being said I believe ISRG is a fundamental short due to not enough long-term growth to justify the valuation given my view that half (systems) of its revenues and profits will be difficult to grow coupled with good odds that the other half (instruments & accessories) does not have as much growth potential as management claims.

Catalyst

Systems sales are weaker than expected leading to lower than expected growth/missed earnings estimates and a subsequent re-rating of ISRG's P/E multiple by the market.