This content is from: Portfolio
CargoMetrics Cracks the Code on Shipping Data
Scott Borgerson and his team of quants at hedge fund firm CargoMetrics are using satellite intel on ships to identify mispriced securities.
One late afternoon last November, as a ping-pong game echoed through the floor at CargoMetrics Technologies Boston office, CEO Scott Borgerson was watching over the shoulder of Arturo Ramos, whos responsible for developing investment strategies with astrophysicist Ronnie Hoogerwerf. At Ramoss feet sat Helios, his brindle pit-bull-and-greyhound mix. All three men were staring at a computer screen, tracking satellite signals from oil tankers sailing through the Strait of Malacca, the choke point between the Indian Ocean and the South China Sea where 40 percent of the worlds cargo trade moves by ship.
CargoMetrics, a start-up investment firm, is not your typical money manager or hedge fund. It was originally set up to supply information on cargo shipping to commodities traders, among others. Now it links satellite signals, historical shipping data and proprietary analytics for its own trading in commodities, currencies and equity index futures.
There was an air of excitement in the office that day because the signals were continuing to show a slowdown in shipping that had earlier triggered the firms automated trading system to short West Texas Intermediate (WTI) oil futures. Two days later the U.S. Department of Energys official report came out, confirming the firms hunch, and the oil futures market reacted accordingly.
We nailed it for our biggest return of the year, says Borgerson, who had reason to breathe more easily. His backers were watching closely. They include Blackstone Alternative Asset Management (BAAM), the worlds largest hedge fund allocator, and seven wealthy tech and business leaders. Among them: former Lotus Development Corp. CEO Jim Manzi, who also had a long career at IBM Corp.
Compelling these investors and Borgerson to pursue the shipping slice of the economy is the simple fact that in this era of globalization 50,000 ships carry 90 percent of the $18.5 trillion in annual world trade.
Thats no secret, of course, but Borgerson and CargoMetrics backers maintain that the firm is well ahead of any other investment manager in harnessing such information for a potential big advantage. Its why Borgerson has kept the firm in stealth mode for years. In its earlier iteration, from 2011 to 2014, CargoMetrics was hidden in a back alley, above a restaurant. Now that hes running an investment firm, Borgerson declines to name his investors unless, like Manzi and BAAM, they are willing to be identified.
My vision is to map historically and in real time whats really going on in economic supply and demand across the planet, says the U.S. Coast Guard veteran, who prides himself and the CargoMetrics team on not being prototypical Wall Streeters. The problem is enormous, but the potential reward is huge.
According to Borgerson, CargoMetrics is building a learning machine that will be able to automatically profit from spotting any publicly traded security that is mispriced, using what he refers to as systematic fundamental macro strategies. He calls the firm a new breed of quantitative investment manager. In unguarded moments he sees himself as the Steve Jobs or Elon Musk of portfolio management.
Though his ambitions may sound audacious, one thing is certain: Borgerson doesnt lack in self-confidence. Over the past six years, he has secretly and painstakingly built a firm heavy in Ph.D.s that can manage a database of hundreds of billions of historical shipping records, conduct trillions of calculations on hundreds of computer servers and systematically execute trades in 28 different commodities and currencies.
For his part, Borgerson seems an unlikely architect of such a serious, ambitious endeavor. Easygoing and fond of joking with his colleagues, he is a hands-off manager who credits CargoMetrics investment prowess to his team. His brand of humor comes through even when hes detailing the series of challenges he had starting the firm. After using the phrase It was hard several times, he pauses and adds, Did I mention it was hard?
Although Borgerson declines to provide any specifics about CargoMetrics portfolio, citing the advice of his lawyers, performance during the three years of live trading apparently has been strong enough to keep his backers confident and his team of physicists, software engineers and mathematicians in place. Hopefully, it wont be too long before we can make a more significant investment, says BAAM CEO J. Tomilson Hill. Former Lotus CEO Manzi is optimistic about the firms prospects: It has an unbelievable edge with its historical data.
CargoMetrics was one of the first maritime data analytics companies to seize the potential of the global Automatic Identification System. Ships transmit AIS signals via very high frequency (VHF) radio to receiver devices on other ships or land. Since 2004, large vessels with gross tonnage of 300 or more are required to flash AIS positioning signals every few seconds to avoid collisions. That allows CargoMetrics to pay satellite companies for access to the signals gleaned from 500 miles above the water. The firm uses historical data to identify cargo and aggregation of cargo flow, and then applies sophisticated analysis of financial market correlations to identify buying and selling opportunities.
Were big-data junkies who could not have founded CargoMetrics without the radical breakthroughs of this golden age of technology, Borgerson says. The revolution in cloud computing has been instrumental. CargoMetrics leverages the Amazon Web Services platform to run its analytics and algorithms on hundreds of computer servers at a fraction of the cost of owning and maintaining the hardware itself.
At his firms headquarters where the lobby displays a series of colored semaphore signal flags that spell out the mathematical equation for the surface area of the earth Borgerson leads the way to his server room. Its the size of a closet; inside, a thick pipe carries all the data traffic and analytic formulas CargoMetrics needs. That computing power alone would have cost $30 million to $40 million, Manzi says.
CargoMetrics is pursuing a modern version of an age-old quest. Think of the Rothschild familys use in the 19th century of carrier pigeons and couriers on horseback to bring news from the Napoleonic Wars to their traders in London, or, in the 1980s, oil trader Marc Richs use of satellite phones and binoculars for relaying oil tanker flow.
Other quant-focused Wall Street firms are latching onto the satellite ship-tracking data. But, Borgerson says, I would bet my life on a stack of Bibles that no one in the world has the shipping database and analytics we have. The reason hes so convinced is that from late 2008 he was an early client of the satellite companies that had begun collecting data received from space and on land to build a large database of all the worlds vessel movements in one place.
Thats what caught Hills eye at Blackstone when he learned of CargoMetrics a few years ago. BAAM now has a managed account with the firm. If anyone else tries to replicate what CargoMetrics has, they will be years behind [Borgerson] on data analytics, Hill says. We know that a number of hedge fund data scientists want his data.
But too much reliance on big data can go wrong, say many academicians. There is a huge amount of hype around big data, observes Willy Shih, a professor of management practice at Harvard Business School. Many people are saying, Let the data speak; we dont need theory or modeling. I argue that even with using new, massively parallel computing systems for modeling and simulation, some forces in nature and the economy are still too big and complex for computers to handle.
Shihs skepticism doesnt go as far as to say the data challenge on global trade is too big a puzzle to solve. When informed of the CargoMetrics approach, he called it very valid and creative. They just have to be careful not to throw away efforts to understand causality.
Another big-data scholar, Massachusetts Institute of Technology professor of electrical engineering and computer science Samuel Madden, also urges caution. What worries me is that models become trusted but then fail, he explains. You have to validate and revalidate.
Borgerson grew up in Southeast Missouri, in a home on Rural Route 5 between Festus and Hematite. His father was a former Marine infantry officer and police official, and his mother a high school French and Spanish teacher. The family traveled 15 miles to Crystal City to attend Grace Presbyterian Church, which was central to young Borgersons upbringing: There he was a youth elder, became an Eagle Scout and received a God and Country Award. The church was across the street from the former home of NBA all-star and U.S. senator Bill Bradley, whose backboard Borgerson used for basketball practice.
When it came to choosing what to do after high school, Borgerson was torn between becoming a Presbyterian minister and accepting an appointment to the U.S. Coast Guard Academy or West Point. He went with the Coast Guard because, he says, the humanitarian mission really appealed to me, and I had never been on a boat before.
At the academy, in New London, Connecticut, Borgerson played NCAA tennis and was also a cutup, racking up demerits for such antics as placing a sailboat on the commandant of cadets front lawn and leading bar patrons in a rendition of Semper Paratus, the schools theme song. Still, he graduated with honors and spent the next four years piloting a 367-foot cutter which seized five tons of cocaine in the Caribbean then captaining a patrol boat that saved 30 lives on search-and-rescue missions. From 2001 to 2003 the Coast Guard sent Borgerson to the Fletcher School at Tufts University to earn his masters of arts in law and diplomacy. While at Tufts he volunteered at a Boston homeless shelter for military veterans and founded a Pet Pals therapy program for senior citizens.
Following graduation, from 2003 to 2006, Borgerson taught U.S. history, foreign policy, political geography and maritime studies at the Coast Guard Academy, and co-founded its Institute for Leadership. While there he would get up at 4:00 each morning to work on his Ph.D. thesis exploring U.S. port cities approaches to foreign policy. He would also travel to Boston to complete his course work at Tufts and meet with his adviser, John Curtis Perry.
Borgersons military allegiance runs deep. One weekend last fall he played football in a service academy alumni game. On another he attended the Army-Navy game. Still militarily fit at age 40, the 6-foot-5 Borgerson works out regularly at an inner-city gym aimed at helping youths find an alternative to gang violence; a few weeks ago he was there boxing with ex-convicts and lifting weights.
Leaving the Coast Guard was a hard decision for Borgerson, resulting in part from his frustration with the military bureaucracys stymieing of his bid to get back to sea for security missions. With his degrees in hand, he applied for a fellowship at the Council on Foreign Relations. During the application process he met Edward Morse, now global head of commodities research at Citigroup. Morse was on the CFR selection committee in 2007 and recommended Borgerson as a fellow.
Morse introduced Borgerson to commodities, and to trading terms like contango and backwardation. Morse himself had, earlier in career, gotten the jump on official oil supply data by hiring planes to take photos of the lid heights of oil tanks in Oklahomas Cushing field.
Working for the CFR in New York reconnected Borgerson with his Missouri roots. Bill Bradleys aunt called the former senator to say: The son of a family who went to our church in Crystal City is in New York. Would you welcome him? Bradley did and would later play a part in Borgersons career development.
While at the CFR, Borgerson became an expert on the melting of the North Pole ice cap, writing numerous published articles on its implications; this led him to co-found, with the president of Iceland, the Arctic Circle, a nonprofit designed to encourage discussion of the future of that region. Borgerson recently spoke to 50 international generals and admirals about the Arctic and is co-drafting a proposal for a treaty between the U.S. and Canada that would help resolve the differences the two countries have in allowing international ship and aircraft travel through the Northwest Passage.
His Arctic research led to an aha moment early in 2008, while he was still with the CFR, on a visit to Singapore and the Strait of Malacca with his Fletcher School classmate Rockford Weitz and their former Ph.D. adviser, Perry. Seeing the mass of ships sailing through the strait, Borgerson and Weitz decided to build a data analytics firm using satellite tracking of ships.
Like many successful entrepreneurs, the two struggled to find financing before reaching out to a network of friends and their contacts. One was Randy Beardsworth, who had sat with Borgerson at a 2007 Coast Guard Academy dinner, where Beardsworth, then the Coast Guards chief of law enforcement in Miami, was the guest speaker. Borgerson made references to history and literature, and I thought, Here is a sharp guy, recalls Beardsworth. We have been friends ever since.
But Borgerson didnt turn to his new friend in his initial fund-raising. He came to me in 2009, after he had been turned down by 17 VCs, was maxed out on his credit card, was married and had a newborn son, says Beardsworth, who was reviewing the Department of Homeland Security as part of the Obama administrations transition team.
Beardsworth came to the rescue, not only committing to invest a small amount but introducing his friend to Doug Doan. A West Point graduate and Washington-based angel investor, Doan took to Borgerson right away. To be honest, it wasnt his idea, it was Scott I invested in, says Doan, who provided $100,000 in capital and introduced Borgerson to a few friends, who added $75,000. Manzi came on board as an investor in 2009, having been asked by Bradley to check out Borgersons plan for a data metrics firm. (Manzi knew Bradley from the late 1990s, when the latter was considering a run for U.S. president.)
With Doan, Doans friends and Manzi as investors, CargoMetrics was finally able to garner its first venture capital commitment in early 2010, from Boston-based Ascent Venture Partners. That gave the start-up the capital it needed to hire a bevy of data scientists to build an analytics platform that it could sell to commodity-trading houses and other commercial users. In 2011, CargoMetrics added Summerhill Venture Partners, a Toronto-based firm with a Boston office, to its investor roster, raising roughly $18 million from venture capital and angels for its data business.
By then Borgerson had already begun to contemplate converting CargoMetrics from an information provider into a money manager; he saw the potential to extract powerful trade signals from its technology rather than share it with other market participants for a fee. Among those he consulted was serial entrepreneur Peter Platzer, a friend of one of CargoMetrics original investors. Platzer, a physicist by training, had spent eight years as a quantitative hedge fund manager at Rohatyn Group and Deutsche Bank before co-founding Spire Global, a San Franciscobased company that uses its own fleet of low-orbit satellites to track shipping, in 2012. We had lengthy conversations on how to set up quant trading systems and how [commodities giant] Cargill had made a similar decision to set up its own in-house hedge fund to trade on the information it was gathering, recalls Platzer.
So Borgerson reset his course. Doan describes the decision as a transformative moment for the CargoMetrics co-founder. The military trains you to be a strategic thinker, Doan explains. Scott had been tactical until then, making small pivots, and like a general who sees the theater of war, he moved into strategic mode.
Borgersons ambition to succeed was in no small part fueled by the early turndowns by many venture capital firms and a fierce determination to best the Wall Street bunch at their own game. Theres a lot that motivates me, including if Im honest I have a big chip on my shoulder to beat the prep school, Ivy League, MBA crowd, he says. Theyre bred to make money, but theyre not smarter than everyone else; they just have more patina and connections. (Bred differently, he spent last Thanksgiving visiting his parents in rural Missouri. After breakfast he and his father were in the woods, shooting assault guns at posters of terrorists, with Gunny, his fathers Anatolian shepherd dog.)
Borgersons plan was not met with enthusiasm from the companys then co-CEO, Weitz. CargoMetrics had been gaining clients and meeting its goals, and was on its way to becoming a successful data service provider. Weitz, who now is president of the Gloucester, Massachusettsbased Institute for Global Maritime Studies and an entrepreneur coach at Tufts Fletcher School, did not return e-mails or phone calls asking for comment. For his part, Borgerson says: A ship cannot have two captains. The company simply matured and evolved into a streamlined management structure with one CEO instead of two.
Eventually, Doan went along with Borgersons plan. We believe in Scott and that shipping holds the no-shit, honest truth of what the economy is doing, he says. But buying out the venture capital firms several years ahead of the usual exit time would require a hefty premium over what they had invested.
Once again Borgersons early supporters played a key role. Manzi, a fellow Fletcher School grad who had mentored Borgerson since the companys early days, put up more money (making CargoMetrics one of his single largest investments) and introduced him to a powerful group of wealthy investors. Separately, the CFRs Morse suggested that Borgerson meet with Daniel Freifeld, founder of Washington-based Callaway Capital Management and a former senior adviser on Eurasian energy at the U.S. Department of State. Impressed by Borgersons intellectual honesty, vigor and more than four years of historical data, Freifeld brought the idea to a billionaire third-party investor, who took his advice and became one of CargoMetrics largest backers. I would not have suggested the investment if CargoMetrics had not done the hard part first, adds Freifeld, declining to name the investor.
A chance encounter in the fall of 2012 gave the CargoMetrics team its first taste of real Wall Street trading. Attending an Arctic Imperative conference in Alaska, Borgerson met the CIO of a large investment firm, whom he declines to name. When Borgerson confided his ambition and that CargoMetrics had developed algorithms to trade on its shipping data once it was legally structured to do so, the CIO suggested CargoMetrics provide the analytical models for a separate portfolio the money manager would trade. Live trading using CargoMetrics models began in December 2012.
Manzi brought in longtime banker Gerald Rosenfeld in 2013 to craft and negotiate the move to make CargoMetrics a limited liability investment firm. Rosenfeld acted in a personal role rather than in his position as vice chairman of Lazard and full-time professor and trustee of the New York University School of Law. The whole process took a year and a half. During that time Blackstone checked in as an investor.
Bradley, now an investment banker, has yet to invest in CargoMetrics, explaining that he is unfamiliar with quantitative investing. But he may eventually invest in Borgersons firm, he says, because we are homeboys. I believe in him and that things are going to work out pausing to add with a smile, based on my vast quant experience, of course.
Borgerson has been in stealth mode since CargoMetrics early days, when he moved the firm from an innovation lab near MIT because the shared space was too open. He is much more forthcoming when boasting of the firms world-class talent. The team includes astrophysicists, mathematicians, former hedge fund quants, electrical engineers, a trade lawyer and software developers. Hoogerwerf, who has a Ph.D. in astrophysics from the Netherlands Leiden University, built distributed technical environments for scientists and engineers at Microsoft Corp. Solomon Todesse, who works on quant investment strategies, was head of asset allocation at State Street Global Advisors. Aquil Abdullah, a team leader in the engineering group, was a software engineer in the high-performance-computing group at Microsoft. And senior investment strategist Charles Freifeld (Daniels father) has 40 years experience in futures and commodities markets, including nine with Boston-based commodity trading adviser firm AlphaMetrics Capital Management.
All were self-made people; none were born with a silver spoon, Borgerson notes. One of his blue-collar-background hires was James (Jess) Scully, who joined as chief operating officer in 2011, after his employer Interactive Supercomputing was acquired by Microsoft.
The team we built treasures team success, which is Scotts motto, Scully says. We want shared resources, one P&L, not How much money did my unit make? Both Scully and Borgerson say CargoMetrics is like the Golden State Warriors, a leading NBA basketball team known for putting aside personal glory and playing as a band of brothers having fun.
Borgerson says he fosters a no-ego policy with lots of play because investment teams are built on trust, and playing together builds trust. Team building at CargoMetrics includes pub crawls, picnics at Borgersons house, poker nights, volunteer work in a soup kitchen for the homeless, Red Sox games and visits to museums.
Trips to the Boston docks or Coast Guard base are intended to remind the CargoMetrics team of the real economy. There are also occasional touch a tanker days. On one visit to a tanker, everyone was amazed that the ship was the size of a city building, Borgerson says. They could smell the salt on the deck, he recalls. Wall Street can lose sight of the real fundamentals in the world. I dont want that to happen here.
Unlike the Rothschilds 200 years ago, only a small percentage of the trades that CargoMetrics makes relate to beating official government data. Most simply are aimed at identifying mispricings in the market, using the firms real-time shipping data and proprietary algorithms.
At a whiteboard in his conference room, Borgerson sketches out CargoMetrics general formula. He draws a maritime matrix of three dynamic data sets: geography (Malacca, Brazil, Australia, China, Europe and the U.S.), metrics (ship counts, cargo mass and volume, ship speed and port congestion) and tradable factors (Brent crude versus WTI, as well as mining equities, commodity macro and Asian economic activity). Using satellite data with hundreds of millions of ship positions, CargoMetrics makes trillions of calculations to determine individual cargoes onboard the ships and then to aggregate the cargo flows and compare them with historical shipping data. All that leads to the final comparisons with historical financial market data to find mispricings. If CargoMetrics observes an appreciable decline in export shipping activity in South Africa, for example, its trading models will determine whether that is a significant early-warning sign by considering that information alongside other factors, such as interest rates. If CargoMetrics believes a decline in the rand is forthcoming, it might short it against a basket of other currencies.
This is like a heat map showing opportunity, Borgerson says, noting that CargoMetrics is not trading physical commodities. We are agnostic on whether to be long or short, and let the computers spot where there is a mispricing and liquidity in the markets. He sums up his simple, but still less than revealing, process by writing on the whiteboard Collect, Compute, Trade.
Borgerson says CargoMetrics is building a systematic approach that will work even when cargo cannot be identified on containerships, for instance. It already knows a large percentage of the daily imports and exports into and out of China and island economies such as Japan and Australia. And although the firm cannot glean from its calculations on satellite AIS data the type of cargo, such as iPhones from China, it can measure total flow, which shows present economic activity. CargoMetrics data scientists are working on linking such activity to the firms data set of the past seven years to measure the evolving global economy. That will lead, Borgerson maintains, to more trades on currencies and equity index futures and, eventually, trades on individual equities.
Uncorrelated is a mantra of Borgerson and his team. Well aware that correlated assets sent the performance of most asset managers, including hedge funds, plunging in the financial crisis, CargoMetrics is determined to come up with an antidote. Careful not to say too much, Borgerson lays out the simple principle that the process starts with placing many bets among uncorrelated strategies in different asset classes, like commodities, currencies and equities.
The goal is diversification, staying as market neutral as possible and remaining sensitive to tail risk in different scenarios. CargoMetrics analytic models help find asset classes that are outliers. Those may include a publicly traded instrument such as oil, another commodity or an equity for which shipping information was a leading indicator during times when other asset classes marched in lockstep. The historical ship data is then blended with this new information to seek opportunities.
Identifying mispriced spreads among different trades within an asset class is another way of avoiding the calamity of correlation. Borgerson says the firms models will find instances where one type of oil should be a short trade and another a long one. The same goes for whole asset classes shorting one that will benefit if virtually all asset prices plunge and buying another that will rise when oil prices gain. Were counting cards with the goal of being right maybe 3 percent more than we are wrong, as a way of making profits during good times and staying afloat during times of sudden, unpredictable but far-reaching events, Borgerson says. The key, he adds, is to know your edge and spread your risk. CargoMetrics uncorrelated approach worked during the dismal first three weeks of this year, says Borgerson. Dialing down risk as volatility in the markets soared, the firm was on track in January to have its best month since it began trading.
To improve the firms models, eight of its data scientists hold a weekly strategy meeting, nicknamed the Shackleton Group after the band of sailors shipwrecked in the Antarctic from 1914 to 1917. Hoogerwerf and Ramos co-lead the group. At one recent meeting they were deciding how much risk, including how much liquidity, there was in a possible strategy; reviewing whether to keep previous strategies; and assigning who would research new ones.
The Shackleton Groups meetings are free-form, with a lot of Ive got an idea interjections that disregard official roles. We hit the restart button a lot, says Ramos, a former director of business intelligence and a quantitative economist at law firm Dewey & LeBoeuf who joined CargoMetrics in late 2010. Thats why our motto is Never lose hope. A bet on oil, related to Russias production, was stopped at the last minute in 2014, when Russia invaded Ukraine. Some currency-trading strategies have been abandoned in theory or after failing. Strategies the Shackleton Group likes are passed on to the firms investment committee of Borgerson, Scully and Ramos for a final decision.
CargoMetrics has a unique set of big-data challenges. Historical shipping patterns may not be as useful in the new global economy now that shipping freight prices are plunging, a sign that trade growth rates may be changing. And analysts point out how hard identifying oil cargo can be in certain locations and instances, even in more-predictable economic times. While it may be easy to say that ships leaving the Middle East Gulf are typically carrying crude oil, knowing the type of crude is sometimes quite difficult, says Paulo Nery, senior director of Europe, Middle East and Asia oil for Genscape, a Louisville, Kentuckybased company that analyzes satellite tracking of ships.
Borgerson maintains his team is well aware of the dangers of data mining and getting swamped by noise. If you run computers hard enough, you can convince yourself of anything, he says. To make sure CargoMetrics algorithms for identifying cargo are valid, the firm spot-checks manifest data filed at ports and imposes statistical confidence checks to guard against spurious correlations.
Getting the jump on official government statistics is likely to become tougher too thanks to the recently formed High-Level Group for the Modernization of Official Statistics. Although the U.S. is not a member, Canada is a key player helping to lead the mostly European nation group (including South Korea) in coming up with a global blueprint for measuring and reporting economic activity.
Reflecting on his journey to Wall Street raising money, hiring employees with different skill sets, making changes to CargoMetrics culture, overcoming legal and regulatory hurdles almost gives Borgerson second thoughts about whether he would do it again. Ive sailed ships through tropical storms, captured cocaine smugglers and testified before Congress [on his Arctic research], he says, but this was the hardest.