I have purposely used the word ideas in the chapter title, because I cannot know your requirements from afar. If I do have the opportunity to assess your requirements and software, we will have hundreds, or close to 1,000 requirements.

Once you have these requirements set out, you are ready for a software vendor selection process. Can you just hand over your requirements and expect the demo to provide the results? You can hand them over, but since they are the same template that seems to be all over the world, you need to take key requirements and ask some questions.

The key to asking those questions is not to check off each requirement, it is to gain perspective on both the simple and challenging requirements you have. Showing up with requirements in your head and a dozen other people, will send your demo off on tangents. Everyone wants to be heard, everyone wants answers today, and there are some who are going to derail the project before it even gets started.

Therefore, I will walk you through a scenario of questions to ask front-, middle- and back-office and provide some perspective. I use colorful word choices to help flush out the essence of the situation, not in hopes that it helps you get the answer, but to help you develop the wisdom to craft your own questions.

Front-Office Capability

I understand that what you want to see first is your position report, or what your view will look like. And, that you want to slice and dice the data anyway you want.

Done! Sold to the guy with the big fat wallet!

You can expect any system will have some canned views, and some slicing and dicing, to an extent. Even adding and re-arranging and saving the view for your desktop are done out-of-the-box. Done, sold, right? Not so fast.

Asking these questions will open the hood and provide you much better insight:

Has your software company’s CEO read Mr. Berley’s book, Energy Trading and Risk Management, especially the end?

1. Until they do, we will reschedule.
2. This is a note to most traders, the attitude from some vendors is extraordinarily the most curious thing you will ever see. Trust me. The egos in the C/ETRM space are huge, but until they act like the CEO I articulated in that first book, reschedule their demo. If you have yet to read my introduction, please go back and read so that you at least have some background on my perspective and approach to dealing with people.

Will you show me how many front-office reports you have? Great, please open them.

1. Can I see both my financial and physical positions in one view? Yes Ma’am, and Yes Sir!
2. Can I slice it any way I choose? You betcha.
3. Yes, will be the answer for any other similar inquiries. Don’t stop now though, keep asking questions.

How many, and which, reports do your other clients use out-of-the-box?

1. For the majority, ranging upward and well in the 90^th percentile, these reports, no matter how you slice, dice and re-arrange, will have to be modified. I know many consultants who all say the same thing: clients rarely ever use out-of-the-box reports.Why? Because the reporting will need to be customized for your needs, and be prepared to pay for them. There are no two trading shops who view the world the same. It is not the software vendors issue, sort of.Use a BI tool – generally not a great idea for front-office on-the-fly reporting needs. Especially with a position report. Even the out-of-the box reports may provide instant position reporting, save options are involved. BI tools are great for historical and accounting reports for internal needs, not usually trading reports; though one could customize some BI tools for front-office capability if so desired
2. Out-of-the box reports are generic and relatively few. Be prepared to sit down and craft your reporting needs, just like everyone else in the process is going to do before the software selection is complete.

 

Can I get real-time position reports? Yes, you can.

1. With real-time valuations. If no options exist and you are ok with yesterday’s closing price, then yes you can. Why this is such a big issue, I don’t know. You should know your position, and the approximate value change when the market moves. Having a report in real time should not be a priority.
2. I rarely gamble, but there is an old adage that you shouldn’t count your money anyway. If you have a disciplined, healthy trading/hedging program, intra-day moves in many markets do not matter.
3. In Power, you may need real-time reports, but anywhere else, it’s not even close to necessary.

 

What valuation tools are embedded into the system?

1. Those that don’t know, tell; those that do, don’t. For the most part, the valuation tools you need are only half-baked.
2. Bolt-on time. Get ready to buy a bolt-on system, here and three other places, especially in Power. If you trade options, it is really time to up your game.
3. Shadow settlement for Power. You can have this if you’re prepared for yet another bolt-on system.
4. Trader marked P&L statements. You’re fired. Score-carding is important, but we are about to spend many millions. If you have a finite budget, a personal scorecard is probably not a must-have. Use the book structures and trader slicing to mirror a scorecard to the extent that it’s reasonable.
5. Credit reporting. You probably want a report that will warn you when you get close to your limit. But if you were close to your limit at the beginning of the day, that should be on alert already. You don’t need your system to do that for you.
6. Trade entry. You might want the system to default the fields you use often. It’s possible, but only if you want to spend your whole budget on it. If you do get it, which you can with some systems, when you are sick or on vacation, nobody will be able to enter trades on your behalf without your logon. Would that work for you?
7. How about any, all and automatic? Sure, just hand me the check book and pre-sign some checks.

Trade capture

it appears all-encompassing, show me!

1. Do not assume all of your financial and physical instruments can be inserted and valued.
2. Send the vendor all of your trade types in advance of the demo. Give them some real data. They don’t trade nor do they have time to see your data and tell the next client about it. They really don’t care unless the trade cannot be captured, which is never an issue in a demo. Get that? Never an issue. The trade can be inserted in a multitude of ways, but that doesn’t mean it will value, much less process or settle, especially for physical products.
3. Will the system allow you to insert all of the fixed and variable costs of your physical plants by unit? If you do have that granular of a need, be prepared to challenge that the system can do it, especially if it is going to be valued as an option. How those costs are captured and then valued, especially in options, will be something you need to see to believe.
4. Foreign currency. Again, every last trade type you engage today needs to be reviewed. Show me its price, in every form and fashion, and settle correctly. Timing differences here can really be an issue, especially if you price in one currency and settle in another. Spot, to forward, daily, weekly, average, or whatever your need, send your trades for the demo.

Book structure and price indexes

1. I have lumped these two together as we search for disaggregation within a portfolio. You will find a varying degree of capability in this area. Ask to see how price indexes, in all of the products you trade, can be disaggregated, not just captured.
2. Book structure. At a high level, you can expect you will need to modify this, as most systems disaggregate in four to six buckets. Anything less and you may be able to use the out-of-the-box functionality. I have seen clients ask for six to ten layers with attributes underneath so they can slice and dice about 50 different ways. This not going to happen without major modification to any software.I understand that more and more companies want to disaggregate down to unit level, capture each fixed and variable component, and do it in a book structure. A fee structure is challenging enough for most systems, but to also have your internal cost structure embedded into the reports, you’ll cross the line of capability for any system without modification.
3. Price indexes. Expect to disaggregate into a variety of components to get to the underlying most liquid products in your market. i.e. location, basis, price, quality. If not, the system is limited or you are in a less liquid market to take your price from Timbuktu back to Henry Hub, PJM, or Cushing.
4. Fee structures: I will also include in this section, for clarity, fee structures for contracts, tiered in fees, quantity, price, location, and grade, can be anywhere from simple to rather challenging. Find your most complex fee-based contract and send it in for the demo.

System performance

1. There is a simple formula to follow here. First, you are not getting intra-day, on-the-fly, reporting with current market prices and instant profit and loss statements. But updated positions, certainly. If you add options, though, you will have no update unless someone runs a valuation and loads some prices with today’s date on the feed.
2. Second, if your book is full of options, and especially assets modeled as options, it can take 15 minutes to process. In power, it can take several hours to run these calculations. Speed is always an issue, and I have mentioned this in Big Data. We will have an ever-growing need for Big Data, and the speed to go with it. Be prepared to take it upon yourself to discern how to parcel out tranches of trades into the database, instead of relying on your software vendor. Just a thought.

I have to pick on the front-office. If you have only been a trader your entire career, congratulations. You have felt little pain, save when you lose money. The rest of the process, and the company, are the real beneficiaries of most of these tools. They will do the heavy lifting, as you will soon learn. If you ever want to lead a trading shop as an executive, I suggest you take the time and make the effort to learn a little about how the rest of the process uses your data, and why when you don’t provide enough detail in the design of the system, it costs you in the long run.

That is the secret – the hidden gem in buying an C/ETRM system. If they are included in this process, which is often an afterthought, the middle- and back-office will gain more than data integrity, they will gain much more data for daily reporting and analysis.

As we move forward, I am intentionally spelling out where issues occur in software selection. I want to help, not just to create realistic expectations, but to get people thinking in terms of asking and discerning the software capability by thinking through these examples and crafting ones that fit your business needs.

Middle-Office Capability

Risk

There is little chance your needs are met here from most C/ETRM vendors. Bolt-on. Next.

Middle-office daily control reporting

These are the folks who ensure all trades are captured, fix trade entry errors, kick-off valuations, show up first in the morning, and ensure the positions and P&L are ready and accurate for traders, leaders, accounting and compliance.

1. Can I embed your trading policy and associated limits into the system? Sure, in many systems.
2. Will it automatically notify someone when a limit is breached? Yes, but it will cost you.

Scheduling

1. Depending on the product, there are a myriad of bolt-on logistics software vendors, and ones even schedulers in their own market have never heard of.
2. I will say that the ETRM systems are making improvements to move products, but depending on how long ago – hopefully five to ten clients ago – they started developing each respective product area, you might get some basic functionality.
3. I would highly suggest providing the specific steps to the software vendor, in advance, and have them demo your data.
4. In markets newer to ETRM, such as LNG or oil, there is pain on the street. As I said earlier, some are taking it for you in advance, so find those recent projects and the software they are using, and you may find a software vendor to put on your short-list.

Credit

Forget it. Actually, many are bringing their calculations into their C/ETRM systems with PFE, CE, PCE, and other data attributes to capture collateral and other credit rating attributes. Again, these will be mostly deterministic.

1. One caveat: The credit arena, as with scheduling, has a niche perspective and it is amazing how many credit modules are being used from a different C/ETRM vendor than the rest of the company. I have seen projects where credit gets scoped away after the project starts because the credit department just won’t accept the software – it happens often. My point is, be aware and manage it to ensure we are using tools with the cost/benefit thought process.
2. Or buy a physical optimization tool.

Back-Office Capability

In this area, which is the big one, we have to discuss a myriad of things to ensure there will be no misunderstandings and that you leverage these tools – and are on the project team from day 1!

I have seen projects where the accounting department is nowhere to be found. This is the hidden gem, and the biggest little secret of purchasing a C/ETRM system: the reporting needs of the back-office are the largest by far.

That means Finance has to fully participate, from scoping to selection and through implementation. Accounting has the biggest data needs, not the front-office. If you are in accounting you are saying, “Thank you, someone has finally heard me!” If you recall, I finished graduate school with a BBA in finance and an MBA. But quietly, every last elective I ever enrolled in was in accounting.

It is the reason, too, that the smartest guys in the room were still there 5 years after I left, and those smart alecks owed me $2.5 million. I know what off-balance sheet financing looks like – how to not only hide it, but take that 25% effective rate of interest oil prepay deal and create $25 million of P&L on top of that crazy loan. Add some more fictitious market-to-market, and you understand why I know when not to hold’em or fold’em, but to run. LOL.

We have to keep this in perspective, though. Just as the front-office has been duped by many a C/ETRM salesperson, so has the back-office. First, C/ETRM systems are not accounting systems, per se. Some do it better than others. With the right system, the shell to have a lot of accounting information is there to help you move data into information and to interface that information with other internal systems.

I would be remiss, though, not to dive deeper.

Chart of accounts

I have seen some scary-looking setups after certain players left the building.

You cannot use a C/ETRM system and expect that it is going to act like a general ledger or subledger, yet some inventory valuation capability is available. You can easily get your summary-level chart of accounts, but you are not supplanting or replacing the accounting/ERP system you currently have.

Settlement

I am sure for most this seems simple unless you’ve worked in this area. The number of attributes needed that don’t make it this far in the process, if ever collected, cause a lot of re-work.

Formatting is probably the second most important issue. Those crystal reports, or whatever tool you use, have to be customized for you, not for the software vendors, so more consulting work is required.

There are some markets where settlement processes are not just daily, but intra-daily. Release a cargo, rail-car, or truck and an invoice is triggered. Many ETRM systems will require someone to manually kick-off a valuation, as the valuations usually are automated for daily runs. You can certainly program many settlement runs, but some folks prefer the manual intra-day method.

As we move the data through settlement, there is data that needs to keep flowing, but not necessarily on the invoice. Ensuring that it does flow behind the scenes, there will be more data attributes, some out-of-the-box and some custom columns. Be prepared to add many columns to ensure data is captured at the beginning.

It should be clearly noted that many firms do not have a standard revenue/expense recognition process. I have seen companies go in and out of unrealized, recognized and realized, intra-month. No two companies are alike.

Tax

This is an area that often requires a bolt-on. Rudimentary capability is generally in place, but taxes on taxes, or the ability to choose from a menu of options – exempt and non-exempt, product to product, jurisdictions, states, countries – can be much more time consuming than you know.

FASB/IFRS reporting

You may think something as innocuous as short- vs. long-term asset and liability reporting is out-of-the-box, but don’t be surprised when it’s not. There aren’t a bunch of CPAs sitting around software vendors.

It may be interesting to know, too, that my definition, and your definition of short vs. long-term can differ as well. I have seen short-term defined as up to 18 months. To me it was a rolling 12 months in my trading books. Therefore, look for changes in accounting pronouncements while selecting systems. They will usually be behind in baking many of these into the system.

They will claim to have FASB 133, 157, 161 or IFRS 9, 32 available in some systems. Get a demo to prove it. Don’t expect that a balance sheet, income statement, or cash-flow statement, in the accounting guru format, is going to exist in these tools.

Regulatory reporting

When I started in this business, regulatory reporting was certainly not going to make the cut for any system. That is why there are some bolt-on systems here, especially for hedge testing.

Be prepared to not have clear designations of hedge levels 1, 2 or 3. It will most likely be required as an extension or re-work to fit your needs. Some define these by trade types or price indexes – whatever the rationale, be prepared to see this issue early.

Hedge effectiveness testing. Yes, the user guide says it is there, but it doesn’t quite work as you may think. Differing approaches, perceptions and applications rein it in.

Throw some of this data at the software demo. You now know why I hope you haven’t gone through that phase. You will want to do it again, I assure you. That cost, not only to you, but the software vendor as well, is worth knowing up front.

Dodd-Frank reporting is even available in several systems. It will be basic, but it does suffice. For some, it is a nice-to-have where spreadsheets, for those who report quarterly and do not have tens of thousands of trades, would suffice. For others, it brings value in connecting to the repository of Neverland. You will probably get the main FERC reporting 556, but not likely FERC 745/55.

Therefore, if you think these systems are doing a great job of keeping up with all of the U.S. to IFRS standards, to FERC, NERC, and state regulations, don’t have high expectations here.

Contracts

It is amazing to me how companies have used this area, from basics to complex trades. If it’s not clear, I have seen trades modeled in contracts due to unique trade language, like cross-commodity switching costs, almost as if they were options. To the client, they were simply contract attributes that needed capturing in some form or fashion.

Certainly, you will find many data attributes available in this area, but I see many businesses who cannot find all of their contract attributes in the system. For example, if you have 350 trucks and you lease those out to move the product, those trucks may represent 350 contracts you need captured in your system.

In what demo does the contract department get represented? Often the accounting department is not involved, so contracts are also usually left out.

IT department

I certainly have to visit this area with some wisdom. You already know every last interface, right? Most do. What most don’t have is what I have said above: the data known. High-level scoping and requirements rarely provide IT with a starting point. IT ends up waiting months and months before the interfaces begin. Those generally take longer than expected.

I have written these technical specifications and they are a back and forth, multiple iteration, multi-month challenge – just to get the specs written, much less programmed. Add logistics and IT will be running around endlessly. Add the other areas from tax to trade entry, with assets modeled at a granular level for costing, and you are definitely going to be underpreparing IT for what is about to be thrown over their wall.

Add the data migration alone and that is overwhelming for many projects. The additional columns that usually come out of nowhere, 90% of the way through a project, and the re-work is just beyond expectation.

Regardless of which group I have mentioned, it always boils down to two things:

1. Reporting needs are not well defined, if at all.
2. Data needs, therefore, are not well defined, either.

They will need to not only manage this part of the project, but also the numerous releases, patches, back-ups, security and the extensions that make your system unique. Yes, these systems often have web services, but all of those extensions are going to be local to your system and cause performance issues as well.

Summary

I have only scratched the surface of functionality that one should explore in selecting a C/ETRM system, or any other class of systems mentioned earlier. I surmise, though, that I have given plenty of perspective to help you think about your business in a way that truly uncovers the key functional areas you’ll need to discern.