- Shopping Bag ( 0 items )
From the Publisher
“the book is a practical guide to good inventory management, the reduction of markdowns, and the achievement of higher gross margins.” — Forbes
It doesn't have to be that way. In The New Science of Retailing, supply chain experts Marshall Fisher and Ananth Raman explain how to use analytics to better ...
It doesn't have to be that way. In The New Science of Retailing, supply chain experts Marshall Fisher and Ananth Raman explain how to use analytics to better manage your inventory for faster turns, fewer discounted offerings, and fatter profit margins.
Featuring case studies of retailing exemplars from around the world, this practical new book shows you how to:
· Mine your sales data to identify "homerun" products you're missing
· Reinvent your forecasting and pricing strategies
· Build end-to-end agility into your supply chain
· Establish incentives that align your supply chain partners behind shared objectives
· Extract maximum value from technologies such as point-of-sale scanners and customer loyalty cards
Highly readable and compelling, The New Science of Retailing is your playbook for turning all that data into a wellspring for new profits and unprecedented efficiency.
Retail Valuation: How Investors Value Product Availability and Inventory Management
For years we have heard from managers responsible for operations and supply chain management that they could not get their CEOs and other senior executives excited about operational issues. A common explanation for the CEOs’ lack of interest in operations was because operational issues did not seem to get attention from investors. Frequently, managers have complained to us that investors in their firms do not pay attention to operational metrics like inventory turns. Moreover, they also complain that quarterly pressure to meet short-term earnings precludes their firms from investing in longer leadtime operational-improvement projects, such as improving store operations and customer service. Clearly, investors’ inability or unwillingness to reward operational improvement or investment in operational improvement could be a barrier to implementing rocket-science in retailing.
The good news for managers seeking to improve operational capabilities is that the situation is changing now. In our experience, Wall Street led by a few fund managers and analysts is becoming increasingly savvier about evaluating operational performance and capabilities. For example, as we argue in this chapter, we see inventory becoming increasingly important to retailer valuation, and we have interacted with some investors who watch retailers’ inventory very closely for early signs of good or bad news. Such attention from investors is likely to translate to attention from senior executives within the company as well and could be a blessing for operations managers who have craved such attention for a long time. On the other hand, some other operating managers and retailers might be unprepared for this attention from senior executives and investors.
This chapter examines the relationship between retailers’ stock-market valuation and inventory management capabilities. The chapter first explores how inventory affects a retailer’s economics (and therefore should affect the retailer stock price) before looking closely at the inventory turns metric. It then identifies how the metric needs to be modified to control for the impact of other operational variables that are correlated with turns. Finally, we provide evidence that valuation is in fact a function of inventory performance once these other variables are controlled for or some missing performance metrics are revealed.
Inventory Management and Valuation
How should a retailer’s inventory level and inventory management capabilities affect its stock market valuation? A look at a typical retailer’s financial statements highlights the importance of inventory to retailer financial performance. Inventory is a significant portion of most retailers’ assets, and the cost of financing and warehousing inventory can be substantial relative to a retailer’s profit. Moreover, inventory has associated markdown and obsolescence costs and the lack of inventory when consumers want to purchase the product can also be expensive. Include the added bankruptcy risk that additional inventory imposes upon the retailer, and it would lead us to conclude that savvy investors and financial analysts should have a good understanding of the relationship between inventory and stock market valuation and incorporate this understanding in valuing retail companies. However, the relationship between a firm’s inventory management performance and its stock market valuation is not well understood and according to some industry observers often ignored even by otherwise careful investors. “Wall Street does not get it [the relationship between a retailer’s stock market valuation and its inventory level],” David Berman, a hedge-fund manager specializing in retail stocks who looks at inventory turns very closely . Numerous retail executives have echoed the sentiment too; some have even bemoaned the fact that the lack of Wall Street attention implies insufficient senior management attention to this problem. This, we argue, is likely to change soon inventory management is going to be a vital piece of retailer valuation in the near future.
Inventory levels on a retailer’s balance sheet -- do not get the weight that one would expect them to get in retailer valuation because investors lack appropriate metrics to reward a firm for managing its inventory well. Retailers should carry an optimum amount of inventory and deviations from this optimum in either direction can be expensive; too little inventory results in additional stockouts and poor customer service while too much inventory leads to additional financing, storage and obsolescence costs. It is hard for an observer external to the firm (such as a financial analyst or an investor) to identify the optimum inventory level for a retailer, let alone to know if a given retailer is at their optimum level. The commonly used metric for evaluating the appropriate level of inventory at a firm namely, inventory turns varies widely across even “similar” firms and over time for a single firm. Moreover, the nature of the variation is not well-understood. Consequently in the absence of knowing the appropriate level of inventory-- it is difficult for investors to evaluate and hence, reward good inventory management performance. Moreover, recent academic research also shows that to reward good inventory management, investors will have to control for the effects of other operational variables (such as gross margins, service levels, and proportion of inventory that is obsolete), some of which can be obtained from public financial statements and others that become apparent only periodically. When investors are able either to control for some of these operational variables, such as gross margin, and when some of these other variables become known (e.g., when a firm marks down inventory), we notice a clear correlation between stock market valuation and inventory levels.
How should Inventory Levels affect a Retailer’s Economics and its Valuation?
Should a retailer’s stock market valuation be a function of its inventory level if investors are able to project future sales and inventory for the retailer perfectly?
To understand the impact of inventory levels on a retailer’s valuation, we can examine the relationship between inventory levels and a firm’s earnings and also how inventory levels can affect the firm’s expected future cash flows. We consider each of these in sequence.
Impact on Earnings: Even a casual glance at a typical retailer’s financial statements shows the importance of inventory productivity to a retailer’s earnings. Not only does inventory impose substantial costs, including obsolescence and markdown, retailers lose sales and gross margins when they do not have the appropriate inventory. Moreover, inventory increases the risk of bankruptcy at a retailer.
We can illustrate the cost of carrying inventory with a simple example. Just as an example, consider a retailer whose inventory (valued at cost) is around 11% of sales . Assuming (conservatively) that the cost of financing and warehousing this inventory amounts to even 10% per year, the inventory carrying cost amounts to 1.1% of sales, substantial even for well-run retailers like Gap and Staples, where profit before taxes are typically between 5 and 10% of sales. In addition to financing and warehousing, the inventory also incurs markdown and obsolescence costs. These costs have been growing relative to sales for most retailers; for US department stores for example, markdowns have risen from roughly 8% of sales in 1970 to roughly 25% of sales in the mid 1990s.
What is often forgotten --in being concerned with the costs of having too much inventory-- is that the costs of insufficient inventory or stockouts can often be much higher. Assume for example that 90% of the consumers looking for a particular product find it in stock while the others (i.e., those that do not find the product in stock), choose not to purchase anything or purchase the product they want at another retailer. By not having inventory of the product when the consumer wanted to purchase it, the retailer loses sales and gross margins (on the lost sales), which would have flowed directly to the bottom line. In our example, if the retailer operated with 50% gross margin (a reasonable estimate for retailers for many segments, including apparel and footwear), the lost gross margin (and also net margin) would be 5% of sales. In our experience working with retailers, the cost of stockouts typically exceeds the cost associated with markdowns and financing and storing inventory.
In addition to the impact on earnings shown above, retailers also face higher risk of bankruptcy when carrying additional inventory. Retailers face a high risk of bankruptcy; the problem arises because they need to build assets (stores and inventory) in anticipation of sales. When sales are smaller than expected, a retailer’s cash flow can become negative quite easily.
Impact on Future Cash Flows: Some professional investors believe that the relationship between a retailer’s inventory levels and future cash flow is “huge.” We illustrate this relationship with a simple numerical example (consisting of a sequence of scenarios) before explaining briefly the views of David Berman, the hedge-fund manager mentioned earlier. Throughout this numerical example, which consists of a sequence of numerical scenarios, we will make an unrealistic assumption that the investor can project the retailer’s future sales and inventory perfectly.
Consider an investor seeking to value two apparel retailers, A and B, whose projected performance is summarized in the following table. Gross margins are expected to be 40% of sales at both retailers; SG&A (selling, general and administrative expenses) will be $180 Million per year. Sales at both firms in the first year are expected to be $500 Million. While firm A is NOT expected to grow or decline in sales, firm B is expected to grow at 3% per year; all of the growth is expected to stem from “comp-store sales growth.” In other words, neither firm is expected to add stores in the next few years. As stated earlier, we assume that the projections are accurate for both retailers.
Financial Metric Retailer A Retailer B
Sales 500M 500M
Gross Margin 40% 40%
Net Margin (current year) 4% ($20 Mn) 4% ($20 MN)
Sales Growth (1 year) 0% 3%
Comp Store Sales Growth (1 year) 0% 3%
Table 1: Scenario 1
To evaluate the two firms, we can use albeit with many simplifying assumptions a discounted cash flow model. In such models, a firm’s value is equivalent to the present value of its future cash flows. For simplicity in analysis and exposition, we ignore taxes, depreciation, and capital expenses. Moreover, we project cash flows for 25 years and assume that at the end of year 25, the terminal value of the firm and the inventory is zero. Finally, we assume that the discounting rate is 10% per year.
Gross Margin% 40%
Year 1 2 25
Sales 500 500 500
GM 200 200 200
SG&A 180 180 180
Net Margin 20 20 20
Cash Flow 20 20
Table 2: Sample Cash Flow Projections
Our discounted cash flow analysis valued retailer A at $182 Million; an annual cash flow of $20 Million for 25 years discounted at 10% per year. Retailer B, on the other hand, was valued at $625 Million. The growth rate has a substantial and favorable impact on cash flow because the gross margin from the additional sale accrues to the company’s bottom-line. In year 2, for example, firm B has sales of $515 Million, gross margins of $206 Million, and net earnings and cash flow of $26 Million. This example highlights the importance of comp-store sales growth, and illustrates why analysts often pay close attention to a retailer’s projected growth.
Now, assume that in addition to the data shown above, we discover that Retailer A will operate with 182.5 days of inventory while retailer B will operate with 365 days of inventory.
Retailer A Retailer B
Days of Inventory 182.5 days ($150 MN) 365 days ($300 MN)
Table 3: Scenario 2
Inventory level does not affect retailer A’s cash flow or valuation. Retailer B’s cash flow is affected adversely compared to scenario 1; as the retailer grows, it has to devote part of its cash flow to fund the inventory needed for growth. In year 2 for example, retailer B’s inventory grows from $300 Million to $309 Million; cash flows from the firm are consequently reduced by $9 Million, and now amount to $17 Million (as opposed to $26 Million in scenario 1). Consequently, retailer B’s valuation in scenario 2 is only $578 Million.
In this scenario, retailer B’s inventory is expected to grow at 6% per year, which is higher than the 3% expected growth in sales.
Retailer A Retailer B
Days of Inventory 182.5 days ($150 MN) 365 days ($300 MN)
Inventory Growth Rate 0% 6%
Table 4: Scenario 3
In our discounted cash flow model, the growth in inventory affects future cash flows substantially. In year 2 for example, inventory increases to $318 Million (from $300 Million in year 1), resulting in cash flow being reduced to $8 Million (from $26 Million in scenario 1 and $17 Million in scenario 2). The sharp reduction in cash flow results in firm B having a valuation of $430 Million.
What would the impact be on valuation if retailer B was also expected to write off some obsolete inventory every year? For example, what would the consequences be if retailer B was expected to write off $10 Million of inventory every year?
Retailer A Retailer B
Days of Inventory 182.5 days ($150 MN) 365 days ($300 MN)
Inventory Growth Rate 0% 6%
Inventory Write off per year 0 $10 Million
Table 5: Scenario 4
Clearly, the inventory write off expected every year would reduce expected cash flow by a corresponding amount. Not surprisingly, our discounted cash flow model values retailer B at $339 Million.
Notice that the four scenarios listed above demonstrate how inventory levels (in scenario 2), inventory growth rate (in scenario 3), inventory write-off (in scenario 4) collectively impact a firm’s cash flow and implied valuation.
The logic embedded in the scenarios described above also summarizes the views of David Berman. The relationship between inventory levels and a firm’s valuation, Berman observes, “is ASTOUNDINGLY powerful, but surprisingly few understand why.” Berman’s arguments for the “powerful” relationship between stock price and inventory levels are two-fold: One, Berman notes that rising inventory levels are often a function of the retailer failing to take markdowns in a disciplined way. Berman argues, that this “game” [of failing to take markdowns on obsolete inventory] cannot be played indefinitely and ultimately the “music has to stop.” In other words, Berman is arguing that rising inventory levels are a predictor of likely earnings declines in subsequent years.
Two, Berman argues that sales growth achieved through increasing inventory are probably not as sustainable as growth achieved by greater consumer acceptance of the brand. Investment analysts who typically pay very close attention to “same store sales growth”—fail to distinguish between these two types of growth and hence, over-value a retailer that is increasing sales by adding inventory. Consider a retailer that has historically suffered from stockouts due to insufficient inventory. When the company adds inventory to its stores, sales will go up. However, Berman notes that such sales growth is less sustainable than growth driven by greater consumer acceptance of the brand. Since analysts incorrectly fail to distinguish the two types of sales growth, they end up over-valuing retailers who increase sales with additional inventory.
How in practice -- do Inventory Levels affect a Retailer’s Valuation?
How closely are inventory levels tracked in analyst reports?
Descriptively, the relationship between valuation and inventory levels does not appear as strong as our normative discussion in the previous section would imply. David Berman notes, “Wall Street basically ignores inventory. It’s actually quite amazing to me!” a sentiment echoed by many retail executives as well.
A preliminary analysis of retail investment analyst reports supports this view. Analysts’ attention to a retailer’s inventory level seems to classify retailers under two categories those retailers that have not had inventory problems in the recent past and those that did.
At retailers that had not had substantial inventory problems in the recent past, inventory levels or inventory turns barely get mentioned in analyst reports. A scan of roughly 75 analyst reports for Best Buy (NYSE:BBY), a consumer electronics retailer, shows that analysts largely ignored inventory levels; the words “inventory” or “inventories” was mentioned at a rate of 0.15 times per page. In contrast, the words “sale,” “sales” or “revenue” were mentioned at a rate of 1.83 times per page (or roughly 12 times as often as inventory and inventories).
Inventory does appear to get more attention at retailers that have not had substantial inventory problems in the recent past but are prone to markdowns and discounts (e.g. specialty apparel retailers and department stores with substantial apparel and footwear sales). While analyzing retailers like Abercrombie and Fitch (NYSE:ANF), analysts use informal techniques such as store observations to get a sense of the likelihood of future markdowns. One analyst’s report for example, mentions the analyst noting “a few back tables and racks of clearance tees and tops marked down at 20-40%.”
Analysts appear to pay very close attention to inventory levels once substantial inventory problems come to light at a retailer. Consider the case of Joseph A. Bank (NYSE:JOSB), a men’s clothing retailer, where inventory turns had been dropping steadily during 2004-05 even while sales and gross margin were increasing; in other words, inventory was growing faster than sales. The inventory levels at the company at roughly 350 days were roughly double the inventory levels at competitors like Men’s Wearhouse (NYSE: MW). On June 7th, 2006, -- after close of trading—the company announced its quarterly earnings. The results fell far short of Wall Street's expectations of 46 cents a share. Earnings declined, even as sales rose 18%, as “discounting and rising expenses clipped margins.” Moreover, the retailer’s inventory levels had grown faster than its sales levels even in the most recent quarter in spite of the higher “discounting” that the company had talked about. Not surprisingly, after June 2006, analysts began to pay very close attention to Joseph A. Bank’s inventory levels. Consider the following from a report on September 8, 2006, “On the positive side...Inventory increased only 12% year over year, well below the 20.8% sales growth. This is a significant reduction over 1Q06 inventory growth of 27%, and 38% inventory growth at the end of 4Q05.” Such quantitative analysis was often accompanied by qualitative observations as well. For example, “RISKS: Inventory management - the company carries a higher level of inventory than many of the company's competitors. Because the company has a lower level of fashion risk, in that most of the product offering is very classically styled and has a longer shelf life, we expect the company to carry more inventory than other more fashion sensitive retailers. We also note that the company is building inventory in preparation of new store openings.” In fact, analysis of roughly 200 analyst reports (from mid 2005- December 2007) for JOSB reveals that inventory or inventories was mentioned at the rate of 0.55 times per page while sale, sales or revenues were mentioned 2.28 times per page (or roughly 4 times as often). The importance given to inventory in Joseph A. Bank’s annual reports stands in stark contrast to the Best Buy example mentioned earlier and also other direct competitors like Men’s Wearhouse, where inventory and inventories were mentioned at roughly 0.24 times per page while sale, sales, and revenues were mentioned at the rate of 1.95 times per page (or roughly 8 times as often).
The comparison of inventory-related and salesrelated word counts for Joseph A Bank, Best Buy, and Men’s Wearhouse are summarized in table 6 below.
Joseph A. Bank Best Buy Men’s Wearhouse
“Inventory” and “Inventories” per page 0.55 0.15 0.24
Sale, sales and revenues per page 2.28 1.83 1.95
Ratio of sales count to inventory counts 4.1 12.2 8.1
Table 6: “Inventory” and “Sales” occurrences in Analyst Reports
Inventory Turns: A Commonly Used Benchmark for Evaluating a Retailer’s Inventory Level
How can we assess if a firm has too much or too little inventory?
How should we compare inventory levels across “similar” retailers? Across “dissimilar” retailers? For a single retailer over time?
Retailers carry inventory for many different reasons examples include to leverage scale economies in purchasing and transportation, to buffer against demand and supply uncertainty, and in some cases to stimulate demand through “display inventory.” While carrying too much inventory can drive up costs and the risk of bankruptcy, insufficient inventory can be expensive as well. Thus, the challenge for retailers is to carry the “right” level of inventory.
Multiple factors make it difficult for analysts to determine if a retailer has the “right” level of inventory.
One, to determine if a retailer is carrying too much or too little inventory, an external observer (such as a financial analyst) needs to know not only the retailer’s inventory level but also the retailer’s service level (i.e., the level of product availability) . Higher target service levels usually require a retailer to carry more inventory. If service level could be observed externally, an analyst could get a sense for if the additional inventory level at a retailer was being used to offer better availability to consumers or simply covering up for inefficiency at the retailer. For example, it is possible that a substantial portion of inventory in the retailer’s balance sheet is actually obsolete; this portion would not enhance service level at the retailer. In other words, the analyst cannot see the mix of inventory at the retailer and hence, cannot identify if the retailer is on or below the “inventory-service frontier.” However, service level is never reported in public financial statements (and only rarely tracked even internally), thus rendering it very difficult for analysts to determine if a retailer has too much inventory for a given service level.
Two, external observers (such as investment analysts) and even retail executives often use inventory turns as a benchmark to determine the appropriate level of inventory at a retailer . However, it is not straightforward to use inventory turns as a benchmark because it varies substantially across retailers; moreover, inventory turns vary across even “similar” firms and over time for a single retailer.
It is not surprising to most observers of retailing that inventory turns can differ substantially for retailers from different segments. For example, supermarket chains like Kroger Company (NYSE: KR) achieve roughly 14 inventory turns per year while apparel retailers like The Gap (NYSE: GPS) achieve only around 7 inventory turns per year. Most observers correctly point out that comparing inventory turns for these retailers would be difficult given the considerable differences in their business models.
Substantial differences in inventory turns persist even for retailers within the same segment. For example, among consumer electronics retailers, Radio Shack Corporation (NYSE: RSH) turns its inventory less than thrice per year while Best Buy (NYSE: BBY) achieves more than 7 inventory turns per year. Best Buy’s closest competitor, Circuit City achieves just above 5 inventory turns per year. Why do we such vast differences in performance across three consumer electronics retailers?
In addition, inventory turns also vary substantially for a single retailer over time. During the period 1985-2000, The Gap’s inventory turns varied between 3.6 and 6.3, Best Buy’s between 3.8 and 9.1, and Wal-Mart (NYSE: WMT) between 4.9 and 7.2. Moreover the variation was not correlated with time.
What explains the variation in inventory turns? Is it possible that the variation in turns is correlated with variation in other variables that could be obtained from public financial data? Moreover, if we could identify these variables, is it possible that we could derive an alternate metric for inventory productivity that “controlled” for the variation in these variables?
Explaining the Variation in Inventory Turns
Could the variation in inventory turns be explained in part by variation in other variables like gross margin, the investment in non-inventory assets, and “sales surprise” (i.e., whether the sales in a particular year for a specific retailer were higher than expected)?
Can we use the large amount of public financial data available for retail firms to quantify the relationship between inventory turns and other metrics?
Would the relationship also provide an alternate metric for evaluating inventory productivity?
A number of variables gross margins, the investment in non-inventory assets, and “sales surprise” -- one would reasonably expect, would be correlated with inventory turns.
Let us begin with a retailer’s gross margin ; in this section we will use markup, which is defined as the gross margin divided by the cost of sales for a retailer. It is optimal for a retailer to carry more inventory when the markup is higher. For example, at higher markups, a retailer should be (and usually is) willing to carry more inventory to cover for unforeseen surges in demand because higher markups also imply greater costs associated with sales lost due to insufficient inventory. In other words, we would expect gross margin (or markups) and inventory turns to be negatively correlated. Not surprisingly, some retailers that have high gross margins or markups are termed “earns retailers” and often achieve low inventory turns (e.g., in 2005 Radio Shack Corporation had gross margins of roughly 46% of sales and 2.8 inventory turns a year) while those with low markups and high inventory turns are termed “turns retailers” (e.g. for the year ending September 3rd, 2006 Costco Wholesale Corporation (Nasdaq: COST) had gross margins of 12.5% and 11.5 inventory turns a year).
Similarly, it is reasonable to expect that a retailer should achieve higher inventory turns as its “capital intensity” goes up; that is, as it invests more in non-inventory assets, such as warehouses and information technology. For example, investments in information technology are often justified based on the technology’s ability to enable faster inventory turns at a retailer. Ideally, we would have liked to get precise estimates of a retailer’s investments in different types of non-inventory assets; however, these investments are almost never broken down in a retailer’s public financial statements. Hence, instead we use an alternate metric of capital intensity, the fraction of a retailer’s total assets that is represented by its non-inventory assets .
A third driver of inventory turns is “sales surprise,” which we define as the ratio of actual sales to sales forecast in a particular year. If sales are higher than had been forecast previously, sales surprise would be greater than 1. All else remaining equal, a retailer would be expected to achieve higher inventory turns when sales surprise is higher.
Analyzing the Data to Quantify the Relationships: Using Panel Data Sets
Public financial data provide us with a context in which to test our hypothesis and quantify the relationship regarding the variation in inventory turns. In the United States for example, there are a few hundred public retailers for whom we have at least annual data on inventory turns and a number of other financial metrics.
To conduct our research, we constructed a data set of all public retailers in the United States for the period 1985-2000. Our data set consisted of 311 retailers and we had multiple years of data for each retailer. Details of the data set and the analysis can be found in Gaur et al (2005).
While it is not our intent in this book to describe the technical details of our analysis, it would be appropriate to explain why a naïve analysis of the data set would not yield useful insights and could even be misleading.
Consider for example, figure 1, which shows the time trends in average inventory turns for public retailers for the period 1985-2000. In figure 1, average inventory turns are calculated in two ways: the line on top averages inventory turns at all retailers, the line at the bottom averages cost of goods sold and divides it by the average inventory level at all retailers. Based on either of the lines in figure 1, one could conclude (incorrectly, as we will soon see later in this Chapter) that there are no time trends in inventory turns for public retailers in the United States during the period 1985-2000.
A simple -- albeit a contrived—example can be used to illustrate why the kind of analysis shown in figure 1 can be misleading . Consider a market with two firms whose inventory turns vary over time as shown in figure 2. Clearly, the two firms show downward sloping trends in their inventory turns over time. Taking a simple average of the two firms’ inventory turns would yield the line in figure 3, which shows small fluctuations up and down, but no change in inventory over the 10 years. The changing mix of firms makes it hard to gauge true inventory progress.
To address this problem, our analysis does not draw conclusions based on the performance difference across firms. Stated a tad technically, our analysis allows for a firm “fixed effect” and draws conclusions on the relationships between inventory turns and other variables of interest (e.g., gross margins) based on intra-firm variation in inventory turns.
The Findings from the Data
Inventory turns at a retailer are correlated with markup , capital intensity, and sales surprise. Gaur et al (2005) provide a segment-wise analysis of this relationship. For simplicity in discussion, we restrict our attention here to a “pooled model” that does not distinguish among the different retailing segments. According to the pooled model,
log ITit = Fi + ct - 0.2431 log MUit + 0.2502 log CIit + 0.143 log SSit + eit
ITit = Inventory turnover for firm i in year t.
MUit = Markup for firm i in year t
SSit = Sales surprise for firm i in year t
eit = residual in the equation for firm i in year t
log = of a quantity denotes the logarithm to the base 10 for the appropriate quantity
Fi = fixed effect for firm i
ct = fixed effect for year t
In words, the logarithm of inventory turns for a firm is a function of the markup (MU), capital intensity (CI), and sales surprise (SS) in a particular year. In the above equation, a 10% increase in markup would translate into a 2.3% reduction in inventory turns, a 10% increase in capital intensity translates into 2.4% increase in inventory turns, and a 10% increase in sales surprise leads to a 1.4% increase in inventory turns. This equation explains 67% of the variation in inventory turns within a firm. Because of the presence of firm fixed effects Fi, the model explains 98% of the entire variation in the data set.
Managerial Implications from the Analysis: Time Trends and Benchmarks
The analysis also provides insight on time trends from 1985-2000 in a number of variables including inventory turns, capital intensity, and gross margin (or markup). During the period 1985-2000, inventory turns and markup (or gross margin) declined quite clearly while capital intensity went up.
A more interesting trend at least in our minds is the pattern observed in ct from 1985-2000. Recall that ct is a year “fixed effect” and hence, shows changes in inventory turns after controlling for variation in markup, capital intensity, sales surprise, and firm differences (as measured by the firm fixed effect Fi). Hence, ct can be viewed as a metric of “average” inventory productivity for a given year. Figure 4 shows the trend in ct over time; notice that ct has declined almost steadily, suggesting that inventory productivity has declined over time during this period. The decline in inventory productivity does not necessarily imply that retailers’ ability to manage inventory has declined over time from 1985-2000. A number of other factors for example, a more competitive retail environment or some changes in consumer taste can affect inventory productivity as well. We ourselves are unsure of the causes behind the steady decline in inventory productivity during this period.
Controlling for variation in markup, capital intensity, and sales surprise also yields a benchmark for inventory productivity. The alternate metric that Gaur et al (2005) term “Adjusted Inventory Turns” is a better measure of inventory productivity than inventory turns because it accounts for variation in other variables that also affect inventory productivity .
Inventory turns and adjusted inventory turns can offer divergent insights at times. Consider the following example from the Ruddick Corporation (NYSE: RDK). Ruddick is engaged in two primary businesses: it owns and operates Harris Teeter, Inc., a regional chain of supermarkets in seven southeastern states, and also manufactures and distributes thread and technical textiles. Notice from figure 5 that inventory turns for the company showed a very small decrease from 1985-2000 even while capital intensity was rising. Figure 5 however shows that markup was rising concurrently; consequently, adjusted inventory turns were rising during the period under consideration.
Plotting the Future: A Resurgence in the Importance of Inventory Turns
A number of indicators suggest to us that analysts are paying increasing attention to inventory turns, the heightened attention is not only at firms that have had inventory-related problems in the recent past.
The importance of inventory to retail valuation can be gauged from the conversation one of us had with a Wal-Mart executive recently. The executive noted that inventory had assumed “greater importance” in recent years. “It is now woven into the performance metric at every level of the organization,” was his remark. When pressed to explain the reasons for the “greater importance” attached to inventory, he noted that, “inventory has become more important because our shareholders care more about inventory now than they did in the past.” It was genuinely surprising to see a retail supply chain leader like Wal-Mart acknowledge that inventory had become more important in recent years at Wal-Mart because of shareholder pressure. Think of the pressure less efficient supply chains might face from their shareholders!
Wall Street is causing firms to look more carefully at their inventory levels in another way too: by penalizing these firms disproportionately for inventory write-offs. In his “Inventory Signals” paper, Richard Lai examines the impact of “inventory write-off announcements” on a firm’s valuation. His analysis reveals that the resulting loss in valuation after a write-off is substantially higher than the “imputed” valuation loss that should have resulted based on the size of the write-off. In other words, though Wall Street finds it hard to reward good inventory management performance, it seeks to induce retailers to pay closer attention to inventory management by severely penalizing poor inventory performance.
Third, we see investors collecting data on their own that enable them to augment the information available in public financial data. Investors like David Berman have hired “armies of foot soldiers” to visit individual stores and collect data on store inventory, pricing, markdowns, and freshness of inventory. A leading investment house has systematized this process by hiring a bunch of people to visit around 50 stores every month to look at the inventory and discount levels for a selected “basket” of products at each store. These data enable these investors to obtain insights that are not reported directly, and are difficult to obtain from, public financial statements.
Finally, we are witnessing academic research some of which was described in this chapter -- that leads to tools that will help investors extract information from public financial data. Two streams seem especially promising to us and we will summarize each of them very briefly here.
In a recent study, Gaur validated the “Adjusted Inventory Turns” metric (“AIT”) by comparing the performance of portfolios based on AIT and Inventory Turns (“IT”). In his validation, Gaur evaluated the performance of “zero investment portfolios” based on each of these metrics. To create a zero investment portfolio based on AIT, he ranked retailers by AIT, and then created the portfolio by investing in the top-ranked retailers and selling short the bottom-ranked retailers. The ranks and portfolios were updated each year, and the performance of the portfolio was evaluated after a number of years. The AIT-based zero investment portfolio easily beat the market, the IT-based zero investment portfolio did not. A caveat applies in interpreting these results: the above analysis does not (at least as yet) validate AIT-based portfolios as an investment strategy to get above-market returns. In conducting these studies, it is typically assumed (just as Gaur did) that accurate financial data are available immediately after the end of each year in commercial databases like Compustat®. Such an assumption is generally not valid in the “real world”: to create investment strategies based on AIT (or IT), investment managers will have to develop processes to collect data relating to these variables from earnings announcements and corporate filings. This can be weeks and at times even months before the data appear in a user-friendly form such as Compustat®. Creating such processes can be challenging and expensive; the studies so far have not factored the cost of creating such processes.
A second stream of academic work that is directly relevant to the theme of the current chapter links prices, sales, and inventory in a retailer’s financial statements . Clearly, these metrics are closely linked a retailer can boost sales for example by lowering prices or carrying more inventory, on the other hand, higher amounts of inventory on hand can cause a retailer to lower prices to increase sales (in units) and reduce inventory. Managers have recognized the links among these metrics; it is clearly important for those analyzing financial statements to acknowledge these relationships as well. Incorporating this refinement as this stream of research does -- will enhance the ability of those outside the company to extract insights from public financial data. Then, it will be possible to identify and reward good inventory performance more easily.
1 Retail Valuation 9
How Investors Value Product Availability and Inventory Management
2 Assortment Planning 29
Mining Sales Data to Discover "Home Run" Products You Are Missing
3 Product Life Cycle Planning 61
How to Reinvent Forecasting, Inventory Optimization, and Markdown Pricing
4 Flexible Supply Chains 105
How to Design for Greater Agility End to End
5 Goal Alignment 131
Reducing Perverse Incentive Misalignment
6 Store-Level Execution 151
Increasing Sales Through Better Availability of Products and Store Associates
7 Technological Risk 181
How Retailers Should Assess and Manage Emerging Technologies
8 Companywide Implementation 199
Managerial Issues Affecting Implementation
The Way Forward
About the Authors 251