From Hustler to Quant - The Arms Race in Truck BrokeragePosted December 05, 2019 by
You are probably tired of hearing that software changes everything. But we still underestimate it, in part because examples of twenty-year old processes exist alongside the bleeding edge. The logistics sector can look like a choose-your-own-adventure book where the evidence says either tech is revolutionary or its all hype. But remember that electrification of industry took a full century to diffuse in the developed world: so it will be with the rise of software tied together via the internet.
I want to show a particular way that this is playing out for trucking markets, one that makes me excited about the year ahead. My hypothesis is that select companies have tinkered enough with digitizing truck brokerage to allow us to tap into the crazy amount of data that was flowing under its surface. Access to that data, along with some ability to execute on it at scale, changes what lies at the heart of truck brokerage success. Instead of competing on the personality of reps it means competing on algorithms and data to feed them. In short, a switch from hiring hustlers to hiring quants.
To understand where truck brokerage is going, look at the equity trading markets.
Equity trading profits used to be driven by private information, not necessarily insider trading but more about supply & demand availability and execution details. Driven by a few large brokerages, in the 1970s NASDAQ started publishing median pricing on the most traded stocks. This was later enriched to ongoing best bid & offer. Then depth was also reported (how many shares can be bought or sold at various prices). Since the 1970s, there was constant improvement in the transparency of price & volume, along with increasingly reliable execution when someone wanted to buy or sell a stock. Today even hobby investors could open an account with Ameritrade, Robinhood, or Schwab and get real time transparency to best price and volume, and execute a buy or sell of a stock… all with zero commission costs.
When you imagine a stock trader you probably think about fast transactions, mountains of data, a race to extract signal for noise, and a graduate from MIT quietly monitoring their automated trades across multiple computer screens as charts and figures scroll by. All of this is roughly accurate, but only recently. For equity trading, the improvement of transparency and ability to execute a trade put pressure on stock brokers to compete on their ability to process information.
Back to the Future
Truck brokerages at the start of 2020s look like equity traders at the start of the 1970s. If you had an omniscient view of the spot market you would see the same laneway being transacted at wildly different rates simply based on who could find whom, and how good they were at haggling. Vendors like Chainalytics or DAT sell post-hoc anonymized median pricing for some laneways, but no one knows the discrete prices or who struck them. In real time it is opaque where the market is at. Brokers make significant margins mostly on the basis of their private information. The foundational technologies and processes that changed equity trading by making price and demand transparent, along with guaranteeing execution, just are not yet existent in trucking.
In the absence of transparency and simple execution, brokerages staff up with reps of a specific sort. When market prices are obscure, you need a rep who can roughly memorize the rates and especially a rep who can haggle well. When supply is obscure, you need a rep who will doggedly call carrier after carrier to find a truck. And when execution is manual you need a rep who will complete mountains of clerical tasks just to make offers, check incoming truck lists for opportunities, or lock verbal agreements into rate confirmations. A recent survey by FreightWaves confirmed all this in stark terms. Take a look at what brokerages think is the most important resource to have when launching a new business, and what skill they most want in their reps:
I’m particularly struck by the profile results. The least important aspect of a brokerage rep is their education, while hustle ranks above integrity. This was also the profile of equity traders before foundational changes gave the advantage to evaluating over hustling.
Key Changes in 2019
Things are changing though. I would point to two specific events in 2019. First, we saw many announcements of loadboards or web portals for carriers where loads have “match now” pricing that the carrier can accept instantly. My company, along with Convoy and Uber Freight, have been working that way since late 2016. But J.B.Hunt, Trucker Tools, and Truckstop have all embraced the match-now concept recently. It seems obvious that any new loadboard will now be built this way.
Second, at least one major shipper in the USA has created an algorithmic approach to their spot buying. And crucially they are mirrored by developments at Convoy, UberFreight, and other tech-savvy brokers who made their own developments to provide quotations in real time and automatically. So, you have buying & selling systems facing each other, deciding if they want a quote, requesting it, generating it, and then evaluating it… all without people in the loop.
These two changes are improvements in execution and transparency. I say execution here to mean the process of requesting, offering, countering, or accepting truck moves. The transparency improvement is straightforward: it’s simply much clearer what is the supply and demand at various prices. The result is that both sides no longer staff for hustle, but for analytical edge. The gossip is that buying this way significantly reduced the shipper’s trucking rates. That is backed up from other sources: McKinsey’s reports that statistical or analytical models can reduce costs by 3 to 8 percentage points (in an industry with net margins only around 10%). The experience of my customers is even better, at 7 to 12 percentage points of reduction.
Don’t Curb Your Enthusiasm
While 2020 may be a turning point, the future is unevenly distributed. A handful of companies are undergoing a serious transformation towards quantitative brokering, but most are not. I watch the top-tier brokerages closely and most are laggards in this regard. The worst have no automated execution infrastructure at all, i.e. the only way to make offers, receive counters, and lock-in a deal with a carrier is with clerical work by a person on a phone (or email). The companies that have gone through the effort to build digital execution are still making the price determination manually. To my knowledge, the number of brokers who plan and execute their buying algorithmically can be counted on one hand. But what struck me this year was not that the industry had arrived at its quant-trading future, but that everyone I spoke with agreed it was the right destination. That makes 2020 the start of an arms race to get there. And that is an exciting way to start a new year.