Ending Reporting Madness

It’s a couple weeks after the end of the quarter, and you’re about to enter the recurring board meeting/reporting cycle. You walk down the hall to check in with Lisa, who’s been preparing your data at the end of each cycle for the past two years. She’s not at her desk, which seems unusually tidy, so you ask her colleague, Peter, where she is. “She’s on her honeymoon for the next four weeks, remember?” says Peter. Panic sets in. Lisa has been collecting, cleansing, harmonizing, and preparing all portfolio data for the quarterly board meeting since you joined the investment team. She does such a great job, you never bothered to learn how she does it.

The above situation may not be the script of a blockbuster horror movie, but it can make your heart race just the same. The scariest part is this may be a scenario that happens routinely in your organization, and you don’t even notice. To see if this applies to you (hopefully in a less dramatic way) conduct this short inventory—

‍

Indicate all that apply:

• You occasionally copy-paste data from a PDF into Excel.

• You’ve cursed at your computer because a macro/data import broke.

• You’ve had to re-format your data in Excel before importing into PowerPoint.

• You’ve had to change font size, color, or move around labels in your beautiful slide deck.

• You’ve had to start all over because the administrator sent an amended file last night.

• Although you have a team working for you, you still mess around with PowerPoint (and take an immense pleasure at making slides look exactly like you want.

If you’re being honest with yourself, you probably marked at least three items. If you didn’t (and still claim to be honest), we can guarantee that three of the above are happening regularly on your team without you knowing. We’ve seen it multiple times while auditing institutional investor data and analytical processes during client onboarding.

As an institutional investor, you don’t produce returns; you produce decisions, which then manifest themselves as returns. It’s a subtle yet important difference. Anything beyond working on the quality of your decision is not where your time and energy is best spent—including dealing with data coming in and data going out. According to our research (an audit of 50 institutional investor processes), outsourcing the nonessentials can give you and your investment team back 40% of your time, so you can focus on making high-quality decisions.

We’ve also observed that the few organizations that have identified the right problem are often solving it with the wrong solutions. They attempt either: (1) DIY semi-automation, or (2) Off-the-shelf visualization tools.

‍

DIY semi-automation

The do-it-yourself semi-automation strategy is typically associated with your latest hire. The brilliant kid who just graduated, speaks four languages, maneuvers Excel as if Bruce Lee playing ping-pong with nunchucks in place of a paddle (enjoy the video, but just know—it’s fake), and can code in three different languages (including that snake’s name, what was it again? Anaconda, Python? Or was it both?). Well, who cares about the details, right? After all, you’re the boss. You feel at ease and check the “Digitization” item off your to-do list.

Except the kid with the fancy digitalized skills you’ve sent off to work will build an intricate Web of SQL databases, import and export routines, and data manipulation snippets only they will understand. The reality is your success is linked to the success and performance of your team. Oh, and by the way, if you were hoping for an interactive presentation where you could magically swipe through data and display information with moving graphics…yeah, that’s not gonna happen. You’re still going to get your typically dull PDF.

Now that you’re sitting down and looking through the data, this trailing twelve months (TTM) performance looks odd. You wonder how it compares to the previous TTM. Clearly, you can’t change the date range on the PDF…Damn, it’s 6 p.m. and there’s no one around to ask. You could have only asked “the kid” anyway, and he’s long gone. You look through your files for a few minutes, hopelessly searching. Then you start wondering if you saved the report iteration from a year ago. Ha! You have a flash-back, a sudden stroke of genius, and suddenly you remember. Your fingers navigate confidently through that folder, et voilà! Here it is. You open it with a smile and, to your greatest disappointment, you realize returns were compounded arithmetically back then and not geometrically, as you told the Bruce Lee kid to implement about three months ago.

The figures are incomparable, and you’ve abandoned all hope. It’s 7 p.m., your spouse will be mad (you were so absorbed that you forgot to call to say you’ll be coming home a little late), and you’ve just wasted an hour of your life clicking hopelessly.

Oh what the hell, you think, and dismiss the matter with confidence. The kid will be back tomorrow (what’s his name anyway?) and he’ll fix it. You grab your hat on the way out.

While do-it-yourself semi-automation is exciting and may give you a pleasant sense of control, it fails to scale in three major ways:

Current technology has surpassed do-it-yourself methods.

You need distributed computing and dynamic interfaces, neither of which are achieved via a few lines of code developed in-house.

Data management efforts to power it are systematically underestimated.

You can’t comprehend the gargantuan effort it will be to pull all portfolio information into one place yet, let alone the efforts to source, maintain, and attach market data to power analytics.

True cost of ownership is underestimated by a factor of 10.

You’re spending 10 times more on hidden costs (salaries, time, coordination) vs. an outsourced solution whose costs are explicit (we send you an invoice) and only about 1/10 of the internally-built approach will cost you.

‍

Off-the-shelf visualization tools

This next false savior is an extension of the DIY mentality. You’re buying tools like SpotFire, Tableau, Visme, PowerBI, Domo, or Qlik because the charts look cool and hey, they’re relatively cheap! Except for the fact that you’ll soon find a team of implementation consultants squatting in your conference room helping your team set up the reports. Before you know it, you’re over budget and your ETA became a rolling procrastination exercise (mañana gringo). Worst of all, you haven’t got a clue as to what the problem is, let alone how to solve it. As time goes on you start to give up. But suddenly it seems like hope is shining at the end of the tunnel. They just sent you a meeting invite for something called “Adoption Kick-off.” That’s it, we must be done! You feel relieved and can’t wait to see the output. But here’s what you find:

Key market data is missing.

You’ve got the data related to your portfolio, but you can’t attach benchmarks, ESG scores, fundamental data, ownership statistics, or other basic information you’d like to aggregate at the portfolio level. You’re told this would be a “massive endeavor” to pull off, and an expensive one.

Key analytics are missing.

You can perform basic operations on your portfolio, which make sense. Summing the dollar P&L of the stocks held in the portfolio over the period yields the right P&L, but doing so in bps yields weird results. “Why aren’t the bps attribution figures adding up to the total bps P&L of the portfolio?” You’re told that compounding is a non-linear operation and that your business intelligence tool, which was built to measure revenue growth among a pharmaceutical salesforce, can’t cope with the subtle nuance of financial math. And that is just the start. You soon realize you can’t aggregate risk either (VaR aggregation is another non-linear exercise).

Maneuvering is still centralized.

You thought you were going to be the cool, calm, confident investor looking at all the key numbers related to your portfolio on an iPad while occasionally contemplating the skyline out the window. You’re soon disappointed. Any change request – even changing dates on an attribution report – needs to be submitted by a ticket.

You had envisioned the coaching sessions with your junior portfolio manager to be supported by the tool. You had dreamed of getting answers to questions emerging during the meeting right then and there, with a few swipes and clicks. “Oh, you feel like you’ve gotten all your bets wrong? Or you want to know whether it was alpha or beta that detracted for a stock? Well let’s click here to find out.” But no, no, and no. You’re not getting the answers. All you’re getting is a request to submit tickets.

You soon realize that, despite the cool visualizations, every minute request must still go through your reporting team. And the biggest disappointment of all? There has not been any decentralization. You feel like you’re back at square one.

Processing speed is still slow.

You’ve managed to do something about the above, and after a few tickets (you’ve gotten used to calling them that by now) you manage to get the option to change dates in your attribution report by yourself. Except you’ve now been waiting for three painful minutes since you changed the date range. This is because, unbeknownst to you, Peter from the performance team is running a data refresh on the server. You call them up, tell them to drop all requests competing for server capacity, and try again. You still wait for three minutes. You call them once more, your tone of voice has risen ever so slightly. After being reassured that nobody is taking server capacity, you’re told,“Well, this is a complicated analysis anyway, it’ll take about ten minutes.” You’re out of time and being called into the next meeting.

Unstructured data is not used.

You’re adamant about getting a full picture of your external investments that goes beyond the obvious performance report. When you select and invest in a manager, you want to know about their latest public filings, the exposures as reported in their factsheets, the VaR that they publish, and the overlap of securities held with your internal teams managing equities. But unfortunately you can’t get it all in the same system, you’re told “the information is not comparable.” You wonder how Google manages to combine all the data they do (GPS positioning with phone numbers, websites, time stamps, real-time traffic information, etc.) when you can’t even get basic information about a manager.

If the above resonates with you, you’re not alone. We’ve been there before and stepped away from our jobs as investors to make yours more pleasurable. Give us a call, and we’ll help you get into the new millennium of portfolio intelligence.