The coronavirus disease 2019 (COVID-19) is bringing new urgency to the development of freight price forecasting tools that can help shippers quickly adapt to fast-changing market conditions and manage risk that has skyrocketed with the advance of the global pandemic.
Shippers are turning to technology for greater market clarity as uncertainty deepens, but businesses trying to figure out how high or low truck pricing will be six months to a year from now are finding forecasts tougher than ever to nail down. Even forecasts from the beginning of March may now seem unrealistic as the number of COVID-19 cases and deaths in the United States rises and measures aimed at slowing the spread send the US economy into recession.
Primarily, shippers are looking for projections that will help them avoid a budget bonfire like the one that scorched them two years ago, when a spike in economic activity collided with tightening truck capacity. That clash pushed truckload rates up by double-digit percentages in late 2017 and early 2018, rendering previously negotiated rates obsolete.
“What shippers really want to do is ensure they’re paying a fair price,” said Drew McElroy, CEO of Transfix, a digital truckload marketplace that uses its own predictive pricing capability to guarantee shippers spot rates up to one year out. He said his platform’s data is both expanding and becoming much more granular, lane by lane, allowing for more detailed contract bids.
“If you can show them how your forecast reflects the way pricing has evolved in a freight lane over the last six months, all of a sudden the classic argument over cutting 10 cents a mile goes away,” McElroy said. “You start making the pricing conversation about facts instead of emotion.”
But can any pricing forecast survive the surprise pummeling delivered by COVID-19? Forecasters say their predictions will obviously change, but historical trends won’t. They must carefully match and analyze current and historical data and fine tune models and predictions constantly. That’s as true, they say, for a truckload pricing outlook as it is for any economic forecast.
As recently as the first quarter, UPS-owned freight broker Coyote Logistics, which produces the monthly “Coyote Curve” pricing forecast for US truckload rates, still expected spot rates to become “inflationary” — i.e., rising year over year — at some point in 2020. Coyote is recalculating its curve to determine whether it needs to be dragged down further.
There is a growing fear in the market that the unique and unprecedented nature of the COVID-19 pandemic makes its effect on freight flows and rates impossible to predict, but Ken Adamo, chief of analytics at DAT Solutions, threw some cold water on that idea.
“The entire idea that you can’t rely on history to predict the future is preposterous,” Adamo told JOC.com. “Forecasting is an art and a science. Sometimes you rely on one more than the other. In conditions like this, it’s more of an art form.” Pandemics like this one are uncommon, but disasters are not, Adamo points out.
“We can think of hurricanes, port strikes, and the disruption caused by the Hanjin Shipping bankruptcy,” Adamo said. “We know what the impact on the markets was. If we see similarities (with the impact of COVID-19), we can make educated assumptions.” The coronavirus pandemic, he said, “is coming on like a hurricane and sticking around like Hanjin.”
We can also look back to the 2009 recession, and the length of time it took to recover from its shock. In just one example from the freight transportation sector, the 50 largest US truck companies by revenue took two years to recover the $16.7 billion in combined revenue they lost in 2009, according to annual rankings prepared by SJ Consulting Group for JOC.com.
The deep uncertainty and disruption caused by COVID-19 may be accelerating shipper demand for predictive pricing tools, but the underlying need long predates the pandemic. Its origin can be traced to the growing complexity of supply chains, both international and domestic, and more volatile and frequent economic cycles that push freight pricing to fresh highs and lows.
Software tools designed to help businesses manage the risk inherent in transportation pricing have proliferated in the past couple of years, with products being introduced rapidly as improvements in underlying technology, including greater use of application programming interfaces, machine learning, and artificial intelligence, spur development.
Another major driver is the widespread availability of data, said John T. Engstrom, chief strategy officer at FreightWaves, a data company that promotes its futures and forecasting products via its media arm. “The tools themselves without actual data can’t do much,” Engstrom said. “Also, the computational ability of computers and their capacity has grown exponentially. With loads being tendered through digital means, you can capture more data.”
These pricing tools aren’t coming from just one sector of the technology market. Digital marketplaces such as Transfix and Leaf Logistics are involved, but so are established third-party logistics providers (3PLs) like GlobalTranz, data providers such as FreightWaves, and truck spot market load-board operators DAT Solutions and Truckstop.com.
Shippers now may choose from among rate forecasting software, forward contracts, and freight futures, all of which approach long-term price forecasting from different angles.
With a broader array of options comes opportunity for confusion, however, as shippers try to determine what’s at stake and what’s offered. The means of prediction and methods for dealing with variables change from tool to tool, and their purposes differ as well. Truckload price freight futures, for example, are a financial tool, while forward contracts are an operational one.
“The one thing everybody is driving for is to eliminate as much uncertainty as possible,” said Eileen Hart, vice president of marketing and corporate communications for DAT, which in January began testing a rate forecasting tool that draws on the $68 billion in annual spot truckload freight procured on its marketplace. “We’re chipping away at the dozens of pain points in the supply chain.”
There’s also debate over how — or whether — predictive pricing tools will reduce shippers’ traditional reliance on brokers and 3PLs to manage the seasonal variability of rates and trucking capacity in any given period or year. But while digital logistics startups are among those leading the development of predictive pricing tools, traditional brokers and 3PLs are active as well.
“There’s been a lot of talk about hedging and freight futures,” said Jeff Tucker, CEO of Tucker Company Worldwide, a truck freight brokerage and management company. “As a broker, I am a hedge for my shippers. I subsidize my customers when I contract for two years, and act as a hedge. All of us in the industry are adopting new technology almost on a weekly basis.”
“The nature of the shipper relationship with the broker or platform and the depth and utilization of different options is continuing to evolve,” said McElroy. “Shippers’ own experience with these tools is evolving, and there’s still a healthy amount of skepticism out there in general. The question is how do we move them off the transactional mindset to a deeper relationship.”
Hedging, buying contracts to protect against sharp changes in the price of fuel, pork bellies, or truckload rates, is one approach to price forecasting. At the moment, more emphasis is being placed on tools that give shippers pricing data they can build a contract around, or book a truck.
Although futures trading involves price discovery, its purpose is very different from that of rate forecasting products and may appeal to an entirely different customer. Hedging is a financial tool aimed at balancing out financial risk, not an operational tool to reduce that risk, so it may not be as appealing to actual supply chain practitioners as it is to a derivatives investor, for example. Any comparison of futures and other rate forecasting tools may be one of, if not apples and oranges, at least lemons and limes.
FreightWaves, which launched its futures market with partner Nodal Exchange in March 2019, offers both options, having built freight rate forecasting into its SONAR business intelligence platform. Early last year, CEO Craig Fuller set a goal of 1,000 contracts traded on its futures platform during December 2019 with a total value of $1 million. By the end of the year, however, the exchange had 66 open contracts, said Tom Mallon, vice president of the freight futures market. FreightWaves did not disclose their value.
Mallon noted futures markets of any type depend on volatility, and that’s exactly what disappeared as spot truckload rates started a long, slow decline in early 2019 after hitting record highs in 2018.
“Last year was a very tough market. Rates were historically low with very little volatility,” Mallon said. As rates dropped last year, “even the carriers wanted to be long,” meaning they wanted to buy futures, rather than sell them. “They didn’t want to hedge, they wanted to speculate.” He believes that dynamic will change when rates “revert to norm” and head upward.
Trading activity on the futures exchange did pick up in mid-March, he said, with 20 new contracts representing 20,000 miles. They were executed at $1.375 per mile, slightly above the average $1.34 per mile spot price when the market was launched 12 months ago. “Of equal importance is the fact that there are bids and offers going out through August,” said Mallon.
Although trading has been low, “we have bids and offers, so Nodal is able to produce a set of settlement prices daily to present forward pricing curves,” Mallon said. That curve goes out 16 months, “but as you get beyond six months, it gets much harder to get a view of where rates will really be.”
Buying and selling is the primary purpose of any market, but futures markets also provide a forward view of prices, said Mallon, who spent much of his career in the energy and commodities markets. “From my perspective, the data is another reference point for the market,” he said.
FreightWaves hedged its own bets by expanding beyond its futures exchange to SONAR, a platform that collects and analyzes a broad range of data, including pricing information, and includes a freight price forecasting tool. “If we had only done futures, there would be nothing else here,” Fuller said in a FreightWaves corporate video podcast. “Stories evolve. The thing that makes startups successful is a willingness to experiment.”
For predictive pricing, the next evolutionary step may be the integration of pricing data into broader systems that aren’t simply focused on predicting rates. Currently, price forecasting tools are most often being tested and used as standalone systems designed to provide a glimpse of future rates, but the supply chain gravity of customer demand is pulling them toward forward-looking visibility.
“At GlobalTranz, we have a costing model we use to understand what rates and capacity might look like in a given market in a given lane,” said Russ Felker, CTO at the Phoenix-based 3PL. “But that’s still just the first step down the road of data analytics. The coronavirus will put a magnifying glass on this. We need to get to a more holistic view of the supply chain.”
That’s not a reference to visibility into the physical location of shipments or tractors or vessels, but a deeper forward look at all factors affecting supply chain execution and costs. “Visibility into where things are isn’t enough, you need to know what will happen,” said Anshu Prasad, CEO of Leaf Logistics, which provides a platform on which shippers can commit to long-term forward binding rates with trucking companies.
The shipping industry is “really poor in terms of information sources that look forward, so everybody is grasping at straws,” Prasad said. “There’s a need to differentiate between better predictions and guarantees.” Leaf Logistics, which counts both shippers and carriers as customers, is working on a dynamic contracting model that will help optimize contracts for service and cost, Prasad said.
Leaf Logistics lets shippers commit to forward contract rates a week out or a year or more in advance, but the real value isn’t just in gaining a committed rate for a future date, it’s in using forward-looking price data in new ways to drive inefficiencies and lower costs. In the dynamic contracting project, “they’re experimenting with different pricing models,” Prasad said.
“The fundamental change, on the carrier side, is they will internally price a truck per driver per day rather than a rate per mile,” he said. “The predominant pricing model in contracts is rate per mile, but the question is how much revenue can you generate from a driver in a truck per day? Once you know that, you can determine whether you’ll gain more from forward commitment.”
More accurate forecasts of “target” rates also could help determine a price range that could be included in contracts to trigger an increase or decrease in contract pricing as the market changes, without reopening an agreement or pushing a shipper or carrier into the spot market. That would be a new model for an industry that has traditionally relied heavily on annual rate agreements.
“In trucking, you’ve got an endless cycle” of peaks and valleys in pricing, Adamo said. “When rates are high, people buy a lot of trucks, and then rates go into the gutter. Being out in front of that will set businesses apart.”
These broader systems that offer a level of forward visibility are still on today’s drawing boards. Shippers may not be able to avoid cyclical economic risk, but as these systems become available, they may be able to plan well enough to cushion the economic impact of unforeseen disruptions.
“You can’t predict the coronavirus,” said Prasad, “but you can be better prepared to deal with it.”