|Shares Out. (in M):||0||P/E|
|Market Cap (in $M):||831||P/FCF|
|Net Debt (in $M):||0||EBIT||0||0|
I am recommending the purchase of Coinstar Inc. (Symbol CSTR). In addition to dominating the coin counting machine market (the market for which CSTR is best known), the company has a near 50% investment in a company (Redbox) which is dominating the incredibly fast growing self serve DVD kiosk market. CSTR has a few other businesses that fit within their “fourth wall” strategy which include entertainment machines, e-pay, and money wires. However, it is the first two business opportunities that I believe make CSTR a very compelling investment opportunity.
CSTR has a 90% market share of the coin counting market. The economics of this business are very compelling, with higher than expected growth given their strong cash flow generation. Redbox has been growing very quickly and, as will be discussed below, is within 18 months of a likely IPO which should value CSTR’s share of the business at over $240 million. So when you back out the value of the Redbox stake that CSTR owns, I believe you get a very attractive business, in a market that CSTR dominates with 90% share, for a ridiculously low valuation; near 5 x’s normalized free cash flow.
Overall Business Strategy
The individual business lines will be discussed below but I would, very briefly, like to discuss CSTR’s overall strategy. The company is attempting to dominate the products and services that are sold on what is referred to as the “fourth wall” of super markets and retailers such as Wal-mart. The fourth wall is near the exits, past the check out counters. This wall is historically space that is under utilized by super markets. It is where you will find the coin counting machines, gumball machines, kiddie rides, DVD Kiosks, money transfer stations and the like. CSTR has products and services that are placed in this area that historically had been “dead” space for the stores but which CSTR has helped make very profitable floor space for the markets.
Coin Counting Business Normalized
CSTR manages a network of machines that count loose chance that customers have accumulated. After the change is deposited into the CSTR machine, the customer has the option of receiving a cash voucher or a stored value card. If the customer chooses cash value they are charged an 8.9% transaction fee. If the customer chooses a stored value card, the customer pays no transaction fee. But the economics are the same to CSTR as the 8.9% fee is charged to the retailer that is providing the stored value card. The stored value cards can be redeemed at top retailers such as Amazon, Starbucks, Borders, iTunes and many more.
CSTR dominates this market. They have a 90% share. The company currently has an installed base 14,000 coin counting machines. On 22 different occasions they have had customers stop using CSTR’s service in an attempt to implement the service themselves. 21 of these customers have returned to CSTR. (Safeway is the one company that they lost that has not returned.)
Each new machine costs CSTR between $10,000 and $12,000 to install. But after maturity, which occurs at approximately 3 years in service, each machine generates $17,000 to $18,000 in revenue to CSTR with a 30% EBITDA margin.
The company depreciates machines over a five year period but in the history of the company they have not retired any machines. The average machine in the field is between 5-7 years old. They still have machines in the field that were installed 13 years ago, at the beginning of company history. Once installed the machines have a long life and are cash cows.
Unit level economics are very compelling and it would be very difficult for a potential competitor to reach the critical mass that CSTR has already achieved.
In 2008 the company has guided to coin EBITDA of $150mm. Currently the company is spending much money on CapEx. This money is largely being spent to put new units in the field (as they should) because of the great economics of each machine as I discussed above. In the third quarter alone the company spent $64mm on CapEX. However, the lion’s share of this was installing new machines across the different business lines. In fact only $4.7mm of last quarters CapEx is what you would deem as “maintenance CapEx”. If we are to annualize that number we can assume that if CSTR was to stop growing, CapEx would be in the neighborhood of $20mm per year. (Actually the company guides to a number lower than this for maintenance CapEx, but to be conservative we will use the $20mm number.)
Using those normalized numbers, if CSTR was to stop growing, which I am by no means suggesting, the company would generate approximately $130mm per year in free cash flow.
Coin Counting Business – Growth
Part of the bear story on CSTR is that the coin business is slow growth and that older machines are near maturity and have stopped growing. I think the evidence will show that this is not the case. Machines that are 3 to 8 years old are “comping” between 5 and 7.5%. Older machines still continue to comp in the low single digit range. So, their oldest machines are not seeing the declines that some bears had predicted.
More importantly, there is much growth opportunity in the installed base. In 2004 they acquired an entertainment company (gumball machines, skill cranes, kiddie rides, etc) called American Coin Merchandising, Inc. This acquisition has largely been blamed for recent problems that the company has had (more on this below), but with this acquisition they were afforded their largest growth opportunity going forward.
American Coin had a relationship with Wal-Mart; at the time of the acquisition CSTR did not. Since the acquisition the Coin machines have been installed in 400 Wal-Marts. Wal-Mart has 4,000 total stores CSTR has the opportunity to move into a significant number of these stores. Additionally, CSTR has 60,000 total locations where at least one product (coin, entertainment, e*pay etc) is installed and only 7,000 of these have more than one product. This allows for much potential cross selling of their products.
Furthermore, 75% of the people
So, not only are the machines still “comping” positive, there is still much room for unit expansion within existing customers.
Redbox – Background
Redbox manages a network 6,000 DVD kiosks that feature new releases of popular DVD titles. Customers are able to rent these DVD’s for $1 per day. Their purchase is on their credit card and if they keep the movie for a second day, they are charged an additional dollar. This continues as long as they have the movie. If the customer doesn’t return the movie within 25 days, they are charged $25 as a replacement cost; about how much it would cost the customer to purchase the movie themselves. These kiosks are also located on the forth wall of large retailers. This new product is a perfect cross selling opportunity for CSTR.
CSTR originally invested in
Redbox in the beginning of 2005. The
other strategic investor in Redbox is McDonalds. Both currently are 47.3% owners of
The fact of the matter is that Redbox has been much more successful than expected and the value of Redbox is well north of $130mm (more below) and there is no doubt that CSTR will exercise this option. Part of the reason it has been more successful than expected is that CSTR sales people have been very successful at cross selling this product.
As a condition to this option that CSTR has, McDonalds has the right to force a liquidity event if CSTR exercises the option. McDonalds will force this liquidity event once CSTR exercises their option so CSTR will make sure that Redbox is ready for an IPO before they exercise this option. CSTR has only to the end of next year to make this happen, which in turn would lead to McDonalds forcing a liquidity event. So we can expect a Redbox IPO within the next 12 to 18 months.
Redbox – Valuation
Wall Street analysts are expecting Redbox to have 10,200 units (from 6,000 now) at the end of 2008. The total market is somewhere in the neighborhood of 50,000 units and Redbox is clearly in the lead of the competition. Currently installed units are “comping” at over 30% on a year over year basis. Revenue expectations for 2008 are expected to be approximately $300mm with EBITDA margins in the neighborhood of 20%.
In a business growing as quickly as Redbox (both organically and with new installed units), I think a conservative valuation for an IPO would be approximately 8x’s EBITDA. Using the numbers I gave you above, we can estimate that Redbox’s total valuation at IPO would be $460mm.
[ ($300mm Rev * 20% Margin) = $60mm EBITDA * 8 times = $460mm ]
CSTR would own 51% of this so we can conservatively estimate that CSTR’s stake in Redbox at the time of the IPO would be worth approximately $230mm.
As I type CSTR trades for just under $30 per share with 28mm shares outstanding for an equity value of $840mm. The company has $224mm of long term debt and cash of 161mm. (Much of this cash is not available for operational purposes as it is in the machines or being processed, but it is CSTR’s cash none the less.)
Putting these numbers together you get an enterprise value of $903mm.
If we subtract out the value of Redbox which I derived above at $240mm we are left with an enterprise value of $663mm for CSTR’s core businesses.
Using the normalized free cash flow number from above of $130mm for the coin business we see that in essences you get the very good coin business, of which they control 90% of the market, for only 5 times multiple of free cash flow. This seems ridiculously cheap to me for a business that has such good economics that has created a moat from competition. CSTR is a near monopoly business that has seen 21 of 22 customers who have left them return to CSTR because it was impossible for them to manage the business on their own. The company is still “comping” positive even on its oldest machines and has much growth potential from cross selling and deeper customer penetration; all for a 5 times multiple of free cash flow.
With the growth potential in the core coin business I don’t think it is unreasonable to think that the core business trade a 10x’s multiple of free cash. This would imply a share price of over $50.
Value = [
Equity Value = $1,477mm
Share Price = $1,477mm/ 28mm shares
Share Price = $52.75
In this write up I have ignored many of the products in the fourth wall strategy. I have done this largely because I believe that the interesting part of the story lies in the strong cash flow nature of the core business and the implied “kicker” of the RedBox business.
However, it is important to point out that the stock sold off after the third quarter report largely because of disappointing results in their entertainment division. This division does the kiddie rides, the gumball machines, and the skill cranes. The company attributed this weakness to the broader forces that are affecting all retailers in this market; lower consumer spending. Additionally, the Chinese lead paint scare that hurt many toy manufactures over the last 6 months also affected CSTR’s skill crane business. CSTR had no issues with lead paint in any of their toys, but the consumers definitely had a negative sentiment to this type of toy.
This business long term is a 9 to 10% margin business that even in this bad quarter didn’t lose money for CSTR. Additionally, this is the business that opened the door to Wal-mart so it is providing business in other ways.
The other businesses in the four wall strategy are showing the early stages of success. E* pay is profitable and doing well and money transfers, their newest business, is about to break even.
Recently it was announced that CSTR was piloting an automatic coffee machine in partnership with Starbucks that would also be located on the fourth wall. This is in the very early pilot stages obviously, but could provide upside in the future.
Lastly, the company has been piloting change machines in banks. If they company is able to get traction in banks it could make the growth opportunity even bigger.
In a nut shell, a purchase of CSTR gets you an investment in the dominate provider of a niche market with great business fundamentals for 5x’s normalized free cash flow. Additionally, you get an equity kicker in a business that is growing very fast that is showing tangible results that should be recognized through a monetizing even within the next year and a half.
Ultimately, I think CSTR would be a good investment without Redbox. With Redbox I think it becomes an outstanding opportunity to see capital appreciation over the next 12 to 18 months.
|Entry||12/21/2007 05:11 PM|
|I owned this company for a while and wrote it for VIC - love the coin counting business. I ended up selling off my position for a small gain as the company kept making value destroying investments - both CapEx and acquisitions. Assuming you are right about the valuation of Redbox that is the first good use of capital the company has made in the last 5 years other than coin counters.|
What gives you confidence that management won't continue to spend most its free cash flow on value destroying projects in the future? If I knew for sure they'd use all free cash flow to buy back stock, this would be an obvious winner.
The stock price would be much higher today if the company had used its free cash over the past 5 years only for new coin counters, spending the rest on stock buybacks.
|Entry||12/21/2007 05:25 PM|
|Looks like a fairly strong shareholder base is now in place. Shamrock (9.4%) and RS Investments (12.7%) have both talked about this in interviews in Value Investor Insight in the past year, Shamrock in July and RS Investments in March. Links are below. Michael Mcconnell, head guy at Shamrock (an activist fund) had the following to say on the capital allocation front:|
"Weighing on the stock was the fact that they’d made a series of acquisitions that had driven return on invested capital down. The market was unclear about whether they were going to continue to do that, and whether they’d be able to successfully execute on the acquisitions. With some input from shareholders like us, the company announced a year ago that they’d make no further large acquisitions until they proved they could integrate past ones. The second thing we’re still diligently working on is to better align incentive compensation with the resumption of improved returns on capital"
|Subject||Coin FCF, Capex, others|
|Entry||12/24/2007 12:38 PM|
|How is ebitda - maintenance capex an accurate free chash flow proxy on the coin biz? Doesn't the company have debt/cash interest? Wouldn't it quickly be a cash tax payer if that was the only business? Seems like that we significantly alter the valuation...|
How do you get comfortable that maintenance capex is really maintenance capex? I.e., the life of these machines has kept and kept creaping up -- has maintenance capex moved up too?
Why are you backing off cash being processed and cash in machines? Is that ALL really cash to the company and/or does not have some corresponding liability that you are not including?
|Entry||12/24/2007 01:18 PM|
|Finn posted a reply that beat me to the punch to some degree. Though I would agree with you that in the past they have made acquisitions that were not value creating and appeared to be worse businesses that the coin, it is hard to look at the acquisitions in isolation. |
As I mentioned, the American coin acquisition on first glance would appear to be the type of value destroying acquisition. But that acquisition opened the door to them getting walmart for their coin business. So all in all I think you would have to say that they would do the acquisition again.
But I think what Finn pointed out, that activist share holder groups have made management aware of this issue gives me more confidence that they won't continue policies of old.
|Entry||12/24/2007 02:50 PM|
|To your first question about how EBITDA-maintenance CapEx proxies the cash flow:|
The answer to this isn't specific to CSTR (the same argument would hold true for any company) but the thinking to using EBITDA to proxy normalized cash flow is that as a business is growing quickly, the working capital needs to grow just as quickly. This has the effect of reducing cash flow at the current time. EBITDA proxies the earnings power of the company, which when the company isn't growing, should equal cash flow generated.
So EBITDA allows you to look at different companies and compare them independently of how quickly each is growing.
Of course it is an estimate, but certainly a very commonly used estimate of what cash flow would look like in a normalized state.
To your second question about the company having debt. Yes they do have debt, but the company doesn't need to have debt, and EBITDA allows you look at the company independent of its capital structure. (This gets into the whole modigliani-miller theorem for you corporate finance buffs.) So yes they do have debt, but again I am trying to look at this company independently of its capital structure as well as independently of its growth rate. EBITDA works well as a proxy for this.
Again I am trying to look at it as a normalized business and what the cash flow would look like in this state. That is why I used EBITDA-maintance capex as a proxy. I hope that answers your question.
Your next question about the maintenance capex and my comfort levels. The numbers I have given have been provided by management but they do give LOTS of detail about this on their earnings call. Additionally, I can back into similar
numbers because they tell us how many machines are installed and how much each machine costs. This gives us very good visibility into what they spent for growth cap ex, what is left over i have assumed to be maintenance.
Lastly, about the cash that is being processed and why I am "backing off" that number as you say. The reason i used the cautionary language is that the company can't use that cash to buy back stock or make investments because that is cash that is, by the nature of the business, always going to be part of the operations. It isn't in a checking account where the company has access to it. But if they were to close down shop, that cash would be theirs. and to you corresponding liability question, no this money is not already spoken for as restricted cash or something like that.
|Entry||12/26/2007 10:42 AM|
|I hope you don't mind if I ask a few really dumb questions.|
Many banks have started to offer free coin counting service, such as Commerce Bancorp, as either a service to their customers or a way of getting customers into their lobbies. This seems to be in competition with Coinstar, and there is a lot of room to undercut their 9% fee. How does Coinstar keep growing in the face of this?
A year or two ago, many reports came out indicating that Coinstar machines were not accurate, sometimes off as much as 10%, and rarely in the customer's favor. What has the company done to address these concerns?
There is a small independent coin counting business in my area and I've talked to the guy who runs it. He said that he can't get enough business from "walk-in" (i.e. retail) customers to pay the bills (which are minimal, basically the coin counting machine, coin bags, and a retail location the size of a bathroom). To make ends meet he counts coins for local retail businesses, but he said that this is competitive since it's a pure commodity and their banks offer the same service. I always wondered how Coinstar was going to keep making money in this industry charging what seems to be absurdly high fees or something that anyone can do. Thanks in advance for any comments.
|Subject||The penny risk|
|Entry||12/26/2007 03:18 PM|
|Given the rise of zinc and copper prices over the last few years, I think any mention of Coinstar has to mention the risk of the penny being taken out of circulation. Many consumers would no longer convert coins if this happened (though Coinstar would still have a runoff business as there would be demand for their coins from those who want to melt them down).|
For over two years, the cost has been over a penny to produce a new penny. The meltdown value of pennies made before 1983 is currently 2 cents and while those made after 1983 have a value below 1 cent, it has fluctuated between .5 and .8 cents over the past couple years. Historically, coins get discontinued when their meltdown value exceeds the value of the coin.
The best possible outcome of this situation for Coinstar is for zinc prices to fall and/or for the Mint to switch to pennies being made primarily from steel.
On the other hand, there has never been a coin in the U.S. with as little (inflation-adjusted) buying power as the penny today - so even without the meltdown issue there is historical precedent for eliminating the penny.
Has Coinstar discussed plans for how they would handle a penniless U.S.?
|Entry||01/02/2008 07:49 PM|
Thanks for the note and for your work in backing that out. I saw your post and can not still not reconcile it. If machines really triple volume in the first 3 years as you suggest, there is no way that the cannibalization effect outweighs the ramp-up effect, as described by the CFO in the past 2 conference calls.
See quote below:
2007q2 cc – “When we had de-installs, we had very high coin comps, or same-store sales, and the reason, as you took some machines offline, but the volume just went to adjacent machines. If you look at the Coin installs for the last two quarters, we’ve installed 700 new machines over just the last six months. That has the effect of taking a little volume away from existing machines and redistributing it. So, if you look at the balloon of installs to same-store sales, if your installs go up a little higher than you expect, your same-store sales are going to be a little lower. If your installs are a little lower because of de-installs, your same-store sales should be a little higher. So, I just looked at Q1 as the flip side of what happened last year, and the balloon is absolutely working. So, we’re not discouraged by that at all.”
My impression is that the add-ons - brigther screens, coin-to-card conversions - are not removed in the SSS calculation.
|Subject||Finn - Clarification|
|Entry||01/03/2008 12:17 AM|
|Not really a tripling of volume in 1st 3 years - my first year assumed 6 months of operation, not 12. Double that to get a sense of first full year of revenue.|
Sometimes SSS calculations are done in a surprising and confusing way - if I were you I'd call IR and ask how they calculate it. A company I posted recently to VIC (CTRN) calculates SSS by including only stores that have been in operation before January 1 of the prior year - meaning some stores 23 months old are not included in the SSS calculation - and this caused me to have trouble with certain reconciliations - each time the seeming contradiction was explained by the unique method for calculating SSS.
I do remember that when the Safeway deinstalls happened a few years back, SSS were higher as nearby stores picked up some volume.
Of course, all these fine points will become largely irrelevant if U.S. Government discontinues the penny - that would overwhelm all other factors and cause a big setback to Coinstar. If we get a more fiscally responsible U.S. government 11 months from now, this might be one of the cost cutting measures to be implemented.
|Subject||re: Finn - Clarification|
|Entry||01/03/2008 04:08 PM|
|Thanks. I have a call into IR.|
I wish you good luck in your fight against Americans for Common Cents, but I have a hard time imaging we actually gets rid of the penny. It seems like too big a fight for too little gain. See the following. I walked by the other day and the 100 million penny display in Rockefeller Center is actually pretty cool:
It would be fascinating to see CSTR's breakdown by coin. I would think the coin-to-card feature would change it significantly as it reduces the incentive to mine quarters and dimes, but who knows.
|Entry||01/04/2008 02:17 PM|
|Firstly let me apologize for the delay in addressing your questions. A family emergency has unfortunately kept me away from VIC for the last week and a half. Lots of questions have been posed at that time, I will jump in and address some of them now.|
1-your point about the cash not being fully unencumbered is valid. the $91mm of cash being processed needs to be paid out to the retailers who have "fronted" the cash or service to those who have made change. Your point is valid, but I do not believe it changes the overall thesis of argument. Some of that cash is CSTR's, but I now agree that the bulk of it is not.
2 - I agree with you that the coin-to-card makes much more sense to the consumer. I year ago, like you, thought it was very big "vig" that they were charging. With coin to card the value proposition is clearly greater. This is evidenced by the fact that the majority of their change business is onto cards now.
3 - I agree with you that the CFO is very polished and sometimes comes of as being "slick." I have spoken with him on several occasions and in a one-on-one setting he seems to have a very strong understanding of the business with out as much of the "slick" salesman. This however is something that obviously you need to judge yourself.
4-your second question about how same store sales are affected by new machines. I think this can be attributed to cannibalization. New machines reduce the sss of existing units. However, as the installed base grows, this effect should be diminishing.
5-Entertainment was clearly weak in the most recent quarter. Am I worried about it? of course to some extent. But I think the economics of the other businesses outweigh entertainment long term. Additionally, as I pointed out, the entertainment acquisition is what opened the door to walmart. So it difficult to isolate the effects of that transaction.
I will delve much more deeply into your questions in the next few days. i want to give you a brief answer as I realized their had been a lot of chatter about this name in my absence. Again, sorry for the delay. Much more from me shortly.
And good questions. certainly no offense taken by your tone
|Entry||02/08/2008 02:15 PM|
|What did you think of the qtr and Walmart announcement? Seems like a mixed bag. WMT was expected to roll out coin counting, but losing the entertainment installs and taking huge corresponding charge is a negative. Most analysts already have the + baked in but not the -s, no?|