Tracking wine value

For his MBA thesis, WINE reader David Priilaid designed a new way of tracking good (and not so good) value wines. His interesting and provocative take on a tricky issue earned him a distinction - and should stir a healthy debate.Historically, research has noted a tight correlation between price and quality; especially where the latter can be assessed fairly easily. In these terms we are thus used to paying more money for better products and services.

All this said, in the world of wine, this relationship isn't as rigid as one would expect. Month on month, Wine Magazine reveals the weird miss-match between price and quality. The current assessment of the Cabernet class of 2001 is no different. Take for example the Blue Creek priced at a mere R45 and receiving four stars. Compare that with some of the snake oil retailing at well over double that price but with ratings trawling the bottom half of the quality spectrum.

So what's going on here? Are these anomalies the result of poor proofreading? Or maybe the professional wine tasters were bribed? Clearly none of these are true. On the contrary, what we see here is merely another example of an inefficient wine market unable to price wine according to intrinsic quality. As this article will attempt to illustrate, this inefficiency constitutes a massive opportunity for producers, retailers, and, most especially wine drinkers.

THE CONSUMER-SIDE FOCUS
Marketers tell us that pricing can be viewed from two different perspectives. The first sees price correlated with business costs. These costs include: production, promotion, distribution, marketing, research, and the like, and together with profit, they add up to the price set by the seller.

The second perspective sees price from the eyes of the consumer, not the producer. From this perspective, the price paid includes the sum of money, as well as the time and effort spent, the cognitive activity involved, and so forth. The economist, Adam Smith put the matter succinctly when he wrote, "the real price of everything is the toil and trouble of acquiring it."

Much of the popular wine literature appears to focus on the first of these views - the business-side of pricing. An apt example of this emerged in a Wine article by Christian Eedes published in April 2001 entitled "How much is too much?" The heart of this piece asked what would constitute a reasonable sum of money for a "quality" wine? The writer's approach was inherently production-orientated, and provided a comprehensive breakdown of costs incurred in the wine making process. In subsequent editions of Wine, a number of strident letter-writers took Eedes to task for failing to include other cost-driving factors. By the conclusion of all this correspondence, it became clear just how complex and detailed the business side of wine pricing has become.

Part of the cost inventory noted by Eedes, was quality. Price is inherently quality driven, he observed quite correctly. Higher quality wines are generally more capital intensive, and should therefore attract higher prices. Because of this - and in order to push up prices, many local producers were thus focusing much time and effort on improving the quality of their wines. All this being so, these insights had little to say about the consumer-driven aspects of pricing within the wine industry. If anything, the article served merely to drive home the perception that in South Africa (at least), the producers' sway on price caries far more weight than the consumer. As one letter writer noted in the subsequent July 2002 edition of Wine "It is each producer's call to price as he wishes, as he knows what has gone into the bottle".

While this business-side model of pricing does have relevance, (if only for the producers of wine), from my jaundiced perspective, the primacy of the underlying notion of supply-driven pricing - is contentious to say the least. Why should the wine producer dictate the terms of pricing? Globally the wine industry is suffering an on-going glut of wine. If anything, the power should therefore be shifting to the demand-side of the equation, where tools like the Internet and magazines like Wine can equip the wine lover with sufficient information to buy the best value available. The days of kissing frogs should be long over.

And here's the rub: as a consumer I honestly don't care what kind of expensive oak is used, how fancy the pedigree of the winemaker might be - or how many generations the wine farm has been in the family. To the contrary in fact. As a wine drinker my primary concern is what is in the glass. Any other distraction serves merely as noise.

Using this line of argument, my recently completed MBA thesis (at UCT) sought to examine what consumer-driven factors actually do affect the price of red wine. Three cultivars were assessed in this study: Cabernet, Shiraz and Merlot. The methodology borrowed extensively from different fields: finance, statistics and marketing. Ultimately the thesis was awarded a distinction. To this extent, and upon further refinement, UCT Finance Professor, Paul Van Rensburg and I submitted a paper entitled "An Econometric Model for Identifying Value in South African Red Wine" to the International Journal of Wine Marketing. In January this year we were notified that upon peer review by experts in the field, the paper had been accepted without any revisions - and is due for publication later this year. For us the acceptance of this piece of work is something of a coup - not merely because it shows what can be done when trans-disciplinary academic research is conducted - but more especially because we believe we have pulled off something significant: a means to accurately identify, calculate and represent value in the wine market. I have called this valuation model, WineTracker, and should it become common currency, all players in this market-space will be affected. Such is the effect of this kind of disruptive technology.

THE WINE-TRACKER MODEL DESCRIBED
So how exactly does the model work? Crudely put, Wine-Tracker employs a statistical method known as linear regression. In short this method is designed to sift through a number of candidate variables that might explain the price of a wine, so as to derive a "formula" that best describes the relationship. For the sake of this article, only one potentially explanatory variable has been selected: Wine Magazine's ratings of intrinsic quality. Having established the "best fitting" relationship between quality and price, this formula is morphed into one that speaks now of the relationship between price (what you pay for) - and value (which is what you get.) The equation derived here calculates the true value of each star - and this is based on the price-quality coordinates of each cabernet sampled by Wine's panel of experts. On the scattergram that follows click here to view, these values have been mapped against prices allowing for the depiction of a 'Value Frontier' - that area of the scattergram where wines with the maximum value can be located. Similarly a "Rip-off Frontier" has also been identified. For the value-driven consumer, wines on or about this frontier are to be avoided. For what they offer they are too expensive.

Using the methodology discussed, in the case of the 2001 vintage of Cabernet, a one star wine is valued at R42.57, and each star there after moves up in units of R28.54. This does however not mean that a 2001 three star wine is automatically worth R99.65. Noting that consumers often buy wine on the basis of an earlier vintage, the final quality metric is based on a 50/30/20 quality weighting spanning the last three vintages: that is the current 2001 vintage, as well as 2000 and 1999. This is done so as not prejudice a wine maker who has had good previous years - and has now experienced a "dud" vintage. Thelema; whose historical average is close to 4 stars, is a case in point. Its uncharacteristic 2001 2-star rating is thereby saved by the 3.5 star ratings it received in 2000 and 1999. Using this method, Thelema's (50/30/20) weighted average is therefore 2.75, and not 2.

In the instance where tasting results exist for only two of the three years, the gap year receives the average of the existing two. Where a maiden vintage is being sampled, the previous years have received the same quality rating. This has not prejudiced wines with a longer pedigree since by far the majority of the new wines sampled received ratings at 2.5 stars or below.

APPLYING THE WINE-TRACKER MODEL
Off the scattergram's north-west frontier it can be observed that, in order of merit, those cabernets offering the most value are: Blue Creek, Swartland, Bianco, Stony Brook, Fleur du Cap, Cederberg, Eventide, Excelsior, De Meye and Overgaau. These ten wines are numbered on the graph and span a broad spectrum of quality and price. Details pertaining to their relative value appear in tabularised form below. I have used the term "return' to quantify the extent of value. If a wine is priced at R40.00 and is valued at R80.00, the percentage return on the wine is 100% - in other words you achieved double your money's worth. If that same wine is valued at R60.00 then the return is 50%. A return of -50% is registered if that same bottle was valued at only R20.00.

Table 1: High value cabernet wines on the Value Frontier:

What is worth noting from this table is that value is to be achieved across the price continuum. R72.00 is not an insignificant price to pay for a wine. This not withstanding, the wine in question, Overgaau, is valued at R115.35 - offering a substantial 60.2% return on one's purchase. This does pale somewhat in the light of Blue Creek - offering a return on investment of a mind blowing 173.8%. Relative to the market, the Swartland Indalo, at 3.2 stars, is also substantially under-priced. For Swartland this level of value is not uncommon - year on year this unpretentious cooperative uses its substantial economies of scale to produce reasonable quality at reasonable prices.

In the south-east quadrant of the scattergram, on the so-called "Rip-off Frontier", low value cabernets can be identified. These overpriced wines have also been numbered and include Linton Park Reserve and La Motte - at one star each. Further up the price-quality continuum, one observes wines such as Rust en Vrede (R 120.00) and Morgenhof Reserve (R 150.00). While such estates are much admired and respected by the wine cognoscenti, they appear to be overtrading on their good names. Based on blind tests conducted by experts in their profession, these are not exceptional quality wines - and should be charging prices significantly below their current price points.

Most alarmingly however are the so-called icon wines: the Rustenburg Peter Barlow (R 270.00) and the Rudera (R300.00). Stripping out the effects of status and brand recognition that sight-based assessments confer, and therefore concentrating only on the impressive intrinsic qualities of these two wines, the Wine-Tracker model indicates that these are worth less than half of what they are currently charging. The difference is made up in the cachet and status of these icon brands; an aspect of the drinking experience which Wine Magazine's assessments rating do not take into account.

Ultimately therefore it depends on what is important to the wine drinker. Lovers of intrinsic quality should, for now at least, should consider options like Blue Creek, Overgaau and the Fleur du cap unfiltered. On the other hand those who believe that branding is also important need to ask themselves whether this aspect is in itself worth what they are being asked to pay - well over R 150.00 per bottle.

Table 2: Low value cabernet wines on the "Rip-off Frontier":

A final word about the Line of Fair Value depicted on the scattergram. This line represents what it says - wines that offer neither good nor bad value. Wines located on or about this line offer the buyer the assurance that what they are paying is what the wine is really are worth. In theory at least, this is where all wines should be positioned.

IN CONCLUSION

In broad terms, wine quality is the biggest explicator of price. The fact that so many local wines are marketed by means of the 'brag rights' achieved from quality awards, bares testimony to this fact. As the WineTracker demonstrates however, the reality is that there is a level of deviation from the price-quality correlation, and hence a good number of wines that are either under or over-priced relative to their underlying quality.

The benefits of applying the Wine-Tracker valuation model are thus manifold. First, to the producer, it provides an indication of what price the market should bear and by how much the quality of its wine must be increased in order to charge a desired price level. Second, to the retailer, it would similarly allow better wine buying practices. It may also provide input as to where higher pricing may be appropriately applied. Thirdly, and perhaps most importantly, to us as wine drinkers, it allows for both the more efficient utilisation of our wine purchasing budgets and more insightful relationships with the wines we share with our friends and family. To you then, the reader of Wine Magazine, I hope then that you might find this model useful.

David Priilaid is a lecturer in UCT's School of Management Studies, and is a CRM strategy consultant working in association with The Partnership and Centric Consulting. Correspondence may be directed through Wine Magazine.