If you watch a 1990s movie that features a Wall Street trading floor, you’ll see stressed-out jocks with a phone under each ear, scribbling numbers on scraps of paper and shouting angry messages across the room.
Today, although phones and shouting are still part of the stock-trading world, it’s a lot calmer and quieter, and much more electronic.
So, you could be forgiven for thinking that the days of offline deal-making and manual data entry were in the past, visible only on VHS and Betamax.
But not so fast.
Although the automation of physical commodity trading processes has come a long way over the past few years, it is still creeping its way from the margins into the mainstream.
In this post, I compare the way things were with how they should be, given today’s available technology, and the value that’s being created by those who have adopted it.
If you’re reading this and start feeling left behind, my company and I are here to help.
Physical commodity trading has long been a phone and paper game.
Whether today’s traders use ICE Chat, a company phone line, or email, a majority of their deals are still made offline.
From there, the information is recorded into two systems—one each for the buyer and seller—which range from sophisticated ETRM products to custom software to spreadsheets. Most of the time, deal confirmation remains a mystery.
(Aside: How many times do you think the word “spreadsheet” is going to appear in this blog? Write down your guess and I’ll tell you the answer at the end…)
Now the post-trade process can begin.
After the seller transfers custody of the commodity, the buyer must create and schedule a series of logistical steps to move the commodity they’ve purchased from wherever it is being sold to wherever it should be delivered.
This could be as simple as one leg from Point A to Point B but might involve several legs (Point A to Point B to Point C to…) or require delivery to multiple end points (some from Point A to Point B, and some from Point A to Point C).
Many more complex situations are obviously possible.
At Trellis, we focus on the supply chain for natural gas from wellhead to burner tip, so I’m going to talk primarily about supply logistics here.
A scheduler must figure out which pipeline to choose for each leg—based on available capacity, contractual terms, fuel allowance (how many gas molecules are consumed along the way), and other factors—to maximize their ultimate profit margin.
This requires logging in to the website for each pipeline to ascertain available capacity. In some cases, schedulers manage this data and intelligence, once again, in spreadsheets.
Having crunched the numbers—in a spreadsheet, in their head, or on paper—the scheduler is ready to issue orders for moving the commodity, known as nominations.
These must each be submitted through the desired pipeline’s website, then manually recorded in the buyer’s spreadsheet.
At the end of each intra-day gas cycle, the pipelines decide which orders (or parts of orders) they will accept to maximize throughput – a process known as scheduling.
The scheduler must log back into the pipeline websites to retrieve this information, transcribe it into their spreadsheet, and then adjust for any orders that were declined or only partially accepted.
At the end of the day, pipeline measurement data provides the ultimate record of what was transported. Each pipeline follows a rules-based allocation process to decide which shipper gets what slice of the resulting pie – a process known as actualizations.
Once again, this data must be retrieved from the pipeline websites and typed into the spreadsheet, so that accurate amounts can be billed and collected and the final movement of both money and commodities captured in a system of record.
This cycle of planning, nominating, scheduling, and actualizing takes place multiple times each day, and must be performed for numerous orders moving between different locations and on different pipelines.
Given the number of manual data entries I’ve just described, it’s evident that the risk of error is high. Disputes regularly arise between buyers and sellers, delaying payments and tying up working capital during the process of reconciliation.
At Trellis, we recognize that spreadsheets and other manual systems are error-prone and not scalable.
Building and maintaining them depends on tribal knowledge that is easily lost when a key employee retires or moves to a different company.
More importantly, in the Information Age, they are unnecessary.
So, we set out to solve each of the manual steps using application programming interfaces (API), electronic data interchange (EDI), and system integrations.
Then we took things a step further, adding dashboards for planning and margin and positions management, and other functionality that simplifies and automates natural gas trade and transaction management.
The solution has now been available for several years—and gets better with every release.
Buyers and sellers (and their associated traders and schedulers) can move seamlessly from trade to planning, nomination, scheduling, actualization, reconciliation, and settlement within one platform. A complete, automated trade to post-trade solution.
There’s no longer a need to log in to multiple pipeline websites multiple times each day; the data flows automatically to a customizable dashboard where nominations can be submitted automatically or with a single click.
The impact on human performance and quality of life is staggering.
Users report a reduction in scheduling workload of 30-40 percent.
This can translate into cost savings through headcount reduction, reduced turnover as more satisfied team members work fewer hours, redeployment of resources to value-adding tasks, such as customer service, and higher scheduling throughput, which increases profit margins.
It all adds up to a lot of benefits and a lot fewer sleepless nights. No more worrying whether the spreadsheet will break, or someone will fat-finger a significant mistake.
So why hasn’t every commodity trading organization rushed to automate its physical logistics operation?
There are several possible explanations.
To start with, expectations are low. The commodity trading world is slow to change—especially the oil and gas sector. Everyone is in a hurry to go second.
The solution is still quite new, in enterprise software terms. Even some industry insiders are surprised to learn that a complete solution is commercially available.
Their focus has traditionally been on optimizing trading, with scheduling given the bare minimum of attention and resources.
Then there’s the status quo.
Spreadsheets are cheap, but the custom solutions that have been built on top of them are expensive (let’s not debate the sunk cost fallacy here).
For the most part, the labor cost associated with all that manual effort—the 30-40% excess headcount that we just discussed—is ignored. The payroll is already in place, built into the budget.
And, perhaps most importantly, everyone is too busy being busy. Change is unwelcome. It might require retraining, things might stop working, and it might lead to personnel changes.
On a brighter note, there is general acknowledgement that the inefficiencies and inherent risks associated with manual data entry and spreadsheets are problems that must be solved.
People at higher levels—the VPs responsible for trading margins and cost optimization—are listening and, one by one, engaging.
Fast followers are beginning to adopt the technology now that early adopters have blazed the trail and front-line users are helping their peers see the advantages it brings.
There’s still a long way to go before the sector takes full advantage of this first level of process automation.
The more companies adopt the technology, the better it gets for everyone.
Communication improves. There’s greater collaboration between trading and scheduling teams. More deals—and increasingly complex deals—can get done.
As data flows in near real time, traders can take advantage of shorter deal cycles, complete more transactions in a day, and generate more revenue at lower cost.
Our vision stretches beyond transaction management to bringing pre-trade, trade, and post-trade into a single platform—for both commodities and capacity.
We’re actively working towards such a solution, so watch this space!
I’d love to hear your thoughts and ideas at rakesh@trellisenergy.com.
Oh, I almost forgot ... the word spreadsheet (or spreadsheets) appears 11 times in this 1,400-word post (not counting this sentence). They’re everywhere—and they’re inefficient, insecure, and don’t scale. We can do better!