The video version of the deep dive will drop tomorrow. My sincere apologies for the inconveniences this may cause.
Edited by Brian Birnbaum
Up 560X+ since its IPO, the Costco algorithm is the ultimate blueprint for long lasting organisational success.
I’ll never forget Charlie Munger’s reply when asked in his last interview ever why he was so sure Costco was going to be a huge success. To my delight, he simply replied that the parking spots were wider than at other warehouse club retailers. Although the reply may seem trivial at first, Charlie was pointing to an important truth. That Costco was particularly more obsessed with delighting customers than its competitors and that by extension, success was likely over the long term.
I’ve read all of Costco’s shareholder letters from 1998 to 2023 inclusive and every time the company achieves a new financial record, the CEO in question attributes it to Costco adhering to its principles. These principles are relentlessly seeking to deliver more value to its members at a lower real price, while obeying the law, taking good care of stakeholders and reducing costs. Having read these letters it is apparent to me that Costco has yielded a 560X+ return for investors since IPO by vigilantly optimising its culture since and thus maximising the cost savings that can be passed back to consumers.
By faithfully living our mission statement... to bring the highest quality goods to market at the lowest possible price... we have re- warded our shareholders time and again over the past 20 years; and we are committed to rewarding you, our shareholders, for many years to come.
-Jim Senegal, Costco co-founder and CEO in his 2003 letter to shareholders.
Analysing Costco’s history makes it seem like the company’s journey to current levels of success was a straight line, up and to the right. In the 1998 to 2023 period, the company only failed to attain a yearly financial performance record in 2009. Yet, this illusion of retroactive determinism is the result of countless, unpredictable decisions taken by employees to maximise value for members and minimise price over a long period of time. Charlie couldn’t see the future, but he understood early on that Costco’s algorithm was quantum-entangled with a successful outcome.
Beyond the maniacal focus on value, price and costs the other fundamental aspect of Costco’s algorithm is capital efficiency. By focusing on a limited set of SKUs, Costco can more easily maximise volume per warehouse, which has historically been and continues to be nearly twice that of competitors. This affords Costco more traffic per warehouse and thus more opportunities to optimise operations, as evidenced by their rapid inventory turn over.
Costco turns over all of its inventory in around 28 days, meaning that it can pay providers after receiving money from end customers. This is a key enabler of Costco’s asset light nature and a reflection of how appealing the overall business formula is to end customers. In comparison, Walmart turns its inventory over just 8.92 times. In aggregate this equates to Costco making much better use of its assets than Walmart, as demonstrated by the higher asset turn over ratio depicted in the graph below.
The higher the asset turnover ratio, the more efficient a company is at utilising assets to produce sales.
Costco’s striking capital efficiency is best evidenced by its CapEx versus that of Walmart, relative to their respective net income levels. As of Q3 2023, Walmart’s CapEx was 4.6 times that of Costco, even though net income was only 2.2 times larger. Capital efficiency is a lever that multiplies the impact of other positive business dynamics, such as goodwill. If we have two companies with the same level of goodwill with end customers, the one with higher capital efficiency will perform better financially speaking over the long term. In turn, assuming no changes in culture, higher goodwill will grant a company more opportunities to become more capital efficient.
The goodwill generated by Costco’s notable focus on enhancing value per real dollar spent by customers is further enhanced by the constant reinvestment of profits. Costco’s CapEx has been rising steadily since the foundation of the company. And there are numerous instances of Costco sacrificing margins in the short term in order to effectively share economies of scale with customers. One particularly notable example happened in FY2012, when the global economy was in a poor state following the Great Financial Crisis:
Our gross margin (net sales less merchandise costs) as a percent of net sales decreased in fiscal 2012, largely due to our investment in lowering prices, which is consistent with our goal of maintaining price and value leadership. This is what we do...each and every day!
-Craig Jelinek, Costco CEO in his 2012 letter to shareholders.
While Walmart shares economies of scale with customers too, the return on total capital has been higher for Costco since the aforementioned gross margin sacrifice. The metric has historically been higher for Walmart but, as depicted in the graph below, we see a reversal of the scenario in Q4 2013. Per these numbers, it would seem that Costco’s unreasonable focus on lowering the real price of a select number of SKUs was particularly appealing in the aftermath of the crisis and that the effect has continued to ripple into the future.
The above divergence is what’s leading to Costco’s expanding valuation multiple, that is currently keeping many investors in the sidelines. Costco now trades at slightly over 1.4 times sales, when it was trading at around 0.4 times just before the Great Financial Crisis. While Costco began leaning into its efficiency-enhancing digital transformation heavily at the start of the 2010 decade, in my view the reversal of the above dynamic is another instance of Costco “magically” pulling ahead due to its intangible organisational properties, from which more goodwill and more efficiency have naturally continued to emerge.
Costco’s membership renewal rate has slowly increased from 86% in the late nineties to over 90% at the end of fiscal Q4 2024. At first glance the increment seems unimpressive, but it has been achieved while exponentiating the number of operational warehouses and notably increasing their average sales volume. There is perhaps no better visualisation of the Costco algorithm than the matrix depicted below, which was unfortunately shared by Costco for the last time in their 2019 annual report. Per the CEO remarks in the annual letters, a big driver of this matrix has been the “treasure hunt” ambiance that Costco excels at instilling their warehouses with, which ushers customers into the warehouses and increases the size of their carts.
People have fun while shopping at Costco and this is by design.
Costco’s algorithm strikes me as unique and while at the surface this may seem like an old company with limited upside, I believe they’re just getting started 41 years after their foundation. Their superior ability to achieve leading price to quality at scale and anywhere means they still have a tremendous runway ahead, just in expanding into other continents and infilling their traditional markets. Indeed, Costco’s financial performance over the next decade will be dictated by the health of its algorithm, as has been the case for the past four decades. As outlined previously, the algorithm only seems to be improving.
While in the technology space disruption is permanently around the corner, this is not as true for a business like Costco. Being a low cost, brick-and-mortar-based retailer at scale essentially means that for the business to fail, it either has to kill itself (most likely in the form of cultural degradation) or that the economy has to fail beforehand. This is an extraordinary existential risk profile that is rarely found in the tech space for prolonged periods of time. As such, I view Costco as an appealing alternative to an index fund, with likely far higher risk-adjusted returns for foreseeable decades.
Perhaps the most important component of Costco’s algorithm is its succession engine. Costco has had three CEOs since its foundation, with the first one being one of the two co-founders, Jim Sinegal. The two subsequent CEOs have come from within the company, having spent decades working their way up to the top. During each of their respective tenures, the Costco algorithm has only continued to improve. The mechanism maximises the odds that by the time new CEOs step up, the chances of them spoiling the culture are low. In the Q4 2024 earnings call, current CEO Ron Vachris shared how Costco currently thinks about succession:
Succession planning continues to be a key focal point for us, as we're continually working on identifying the future leaders of our company.
In fiscal year 2024, we promoted 95 new warehouse managers. 85% of those promoted, started at Costco as an hourly employee. This promote from within culture and the long-term career it helps to build is core to who we are as a company, community member, and retailer.
Much like the other components of the Costco algorithm, it is hard to predict what the output will be in the future. But if well cared for, the above system can be relied on to produce capable tenants of the Costco algorithm/culture over time. I hadn’t seen a reference to the succession mechanism in any of the annual shareholder letters from 1993 to 2023 inclusive, so I welcome Vachris making emphasis on the matter. Indeed, a major sign of deterioration of the succession engine would be a CEO from outside the company stepping up.
My general observation is that many of today’s top tech companies have borrowed Costco’s algorithm. A company that keeps margins stable (or lower) while sharing economies of scale with customers and growing fast screams excellence. This is true of companies like Tesla, Amazon and Spotify, that have been and continue to a large extent to be misunderstood by the market. Costco’s multi-decade success shows how value can compound in a platform that’s truly obsessed with their users and that is constantly looking for ways to share back efficiency gains with them.
Costco’s membership model shows us the way forward for modern subscription businesses.
Another fundamental component of the Costco algorithm is its members-only model, which converts at an 80-90% rate to the bottom line, generating over $3.5B of profits yearly. Although the membership fees add a great element of buoyancy to Costco’s financials, what’s interesting is how small the fee is in comparison to net sales. Over time, Costco has continuously found new ways to capture wallet share from their members.
Today’s large subscription businesses, like Spotify for example, make money primarily through membership fees. In the case of Spotify, I believe that their ability to constantly find new ways to add value to its subscribers will over time yield a similar scenario, in which most of the revenue comes from lines of business that are non existent today. Costco makes a substantial amount of its profits by selling gasoline, for example. In turn, each of these new lines of business that Costco deploys makes the membership more attractive and drives more traffic to each warehouse.
Specially intriguing is that Costco only increases prices when the additional value delivered eclipses the delta in the fee. This naturally brings me back to Spotify, which runs the exact same playbook. They have recently surprised the market with a series of price hikes, which consumers have been unfazed by. Additionally, Spotify is now expanding margins by deploying additional verticals with much better unit economics, as Costco has done over time. As such, gross margins are now expected to come in over 30% in the coming earnings report.
Meanwhile, the number of total paid members has grown steadily, as you can see in the graph below.
What’s also fascinating is how Spotify has obliterated all competitors against all odds, just like Costco. The Costco algorithm seems to work across industries and indeed, spotting companies that run it early on is an appealing investment strategy. It is the precursor of numerous and seemingly unlikely success stories across industries. I believe that Hims is another instance of this playbook. Hims is sharing economies of scale with its customers, while increasing margins and printing free cash flow outside of the existing healthcare insurance model, which is a true statistical anomaly.
A notable example of Costco deploying new lines of business is its private label brand called Kirkland, which was launched in 1995. As of Q4 2024, it accounts for just under 30% of all sales. Kirkland offers a broad range of products that are meant to be better than national brands and at a 20% discount. While Costco aims to effectively maximise the lifetime value of its suppliers (just as it does with end customers), my view is that over the long term this brand can generate tremendous externalities. Notably, as Kirkland displaces other top brands found in alternative retail surfaces, this could force suppliers to come under Costco’s wing and non-linearly increase the appeal of a membership.
Another recent example is Costco Logistics. In 2020, Costco acquired Innovel Solutions in order to vertically integrate the home appliance delivery and installation capabilities. Following the acquisition, the services were rebranded to Costco Logistics and as of the end of FY 2024, the service had delivered over 4.5 million items, which is up 29% YoY. This very much resembles the growth of Costco’s e-commerce division, which acts as a complementary extension of the brick and mortar operations.
Whichever way I look, I see signs of the Costco algorithm continuing to enhance the company’s earning power. I emerge from this deep dive with an improved sense of how to spot top-level instances of the algorithm across industries, with the firm but loosely held believe that all five companies that I hold in my portfolio are prime examples of it. I believe this is one of the most important deep dives I have ever done and I hope that you have found it just as useful.
Until next time!
⚡ If you enjoyed the post, please feel free to share with friends, drop a like and leave me a comment.
You can also reach me at:
Twitter: @alc2022
LinkedIn: antoniolinaresc
Hi antonio,
My name is bruno from brazil,
Always enjoy your posts,
About costco and their matrix,
They are still posting the tables every year,
You can find in page 21 of the last 10k (published 10/09/2024)
By the way the trend keeps going im the right direction :)