Max Bowie: A Whine About Wine (But Really About Data)

Many readers will know that I’m fond of the odd tipple, especially a nice glass of wine. But I was well into my 20s before I really gave red wine a chance, when my then-publisher introduced me to its delights. So when events like Smith & Wollensky’s Wine Week come up, I get a little excited. 

Recently, a long-standing market and reference data executive and I caught up over a Wine Week lunch—a true pleasure, but with one shortcoming that will doubtless resonate with data professionals: These great-quality wines weren’t all served in the order on the tasting notes. Sometimes a different server would see our glasses empty and pour whatever they had in their hands. One server might call a wine by its name, while another would refer to the same wine by its varietal or appellation. Did you know that Shiraz and Syrah are the same thing? Pinot Gris and Pinot Grigio, too. And sometimes a server would pour a second glass of something we’d already had. Plus, as glasses piled up on the table, who could keep track of which wine was in what order? We left with our heads spinning—and not just from the wine itself.

Out of Sequence

Am I being too picky? Imagine in the data world, if prices arrived in your systems out of sequence: First, you might accidentally trade on a stale price, and second, you might not be able to immediately tell which way the price is moving. In fact, this is exactly why Gain Capital’s GTX currency ECN only distributes non-tradable reference prices over its new GUDP data protocol, since it is wary of UDP Multicast feeds (affectionately known as “spray-and-pray” for their efficient yet unreliable distribution) dropping data packets or delivering them in the wrong sequence. GUDP is one of several new data protocols recently introduced by GTX in an effort to expand access to its data with more flexible delivery mechanisms, and to attract high-frequency traders.

As a currency ECN, GTX also has to deal with another data management challenge that other market data sources, such as exchanges, do not: With each trade being conducted bilaterally with no central counterparty, participants can only do business with those brokers with whom they have an established relationship, and can only see prices from those brokers. Hence, there is no single, standard feed of all quote and trade data from GTX. Instead, each participant receives a different feed, based on their individual permissions—so in these cases, it’s all the more important that data doesn’t get mixed up.

Imagine in the data world, if prices arrived out of sequence: You might trade on a stale price, or not be able to tell which way the price is moving.

To guard against the risks of data arriving out of sequence, or the wrong data inadvertently arriving and bringing with it extra fees or fines, firms and vendors have developed sophisticated methods of tagging data packets so their lifecycle can be tracked and audited. 

The same principles apply to physical assets and traded commodities. For example, New York-based shipping data vendor ClipperData combines customs and port documentation about ships and their cargos with transponder data from those ships in open water to create a dataset of commodities traffic—what commodity (and how much of it) is being shipped from where to where—that can be used to determine supply and to set prices. Now, the vendor is expanding into slightly different territory—both in terms of datasets and client bases—with a new product that monitors oil well production to generate a prediction for US oil production at a national level. So ClipperData knows all about getting the right data to the right place—after all, if data from one well were confused with another producing substantially more or less, its entire projection of consumption could be skewed.

And especially if you feel skewed, don’t mix consumption with transportation in the case of wine—unless it’s public transportation, that is. Just as your data is more likely to end up in the right place if entrusted to data management experts, you’re more likely to end up in the right place, safely, if you entrust your transport to the experts. Safe travels to you and your data!