Competition is strong in the institutional brokerage market, and niche firms must find a way to stand out. Mischler Financial recognized that its institutional clients needed more in a partner than a strong reputation and competitive pricing – they needed access to information. A series of models developed on the Quantrix Qloud platform has created a “slam dunk” edge for the firm.
Mischler Financial is the financial industry’s oldest broker/dealer firm owned and operated by service-disabled veterans. It serves institutional clients including public entities such as cities, counties, and municipalities, and provides direct access to global equities and fixed income markets, new issue underwriting and syndication within the US Equity and Debt capital markets, and asset management for liquid and alternative investment strategies.
“These large institutions operate very large funds – often in excess of hundreds of millions of dollars,” says Jason Klinghoffer, CFA, Director, Mischler Financial. “But many times the fund managers are public servants without access to the tools and resources a Wall Street money manager has.” Recognizing their budgetary limitations, Mischler sought a way to level the playing field for its customers that buy and sell without access to Bloomberg terminals or costly analytics programs.
Klinghoffer joined the firm in 2014 with a mission to create a new set of analytical models to replace the cumbersome investment strategies and offerings spreadsheets that Mischler provided to clients. One requirement: an online platform for direct client access. “No matter how cool your spreadsheet is, at the end of the day, email is a huge bottleneck,” says Klinghoffer. “It’s inefficient, clients can’t easily access the models on demand, and we can’t track whether or not they are using them.”
Originally he planned on developing an internally housed solution from the ground up, but chose Quantrix after discovering the robustness of the cloud-based Quantrix Qloud. “It has its own admin panel, we can assign logins to clients, and we didn’t have to develop a system to host the analytics,” says Klinghoffer, who uses Qloud tracking functionality to track user logins and usage. “We can see who is logging in and who is accessing and using models, all from within Quantrix. This creates great efficiency in managing our clients’ information and analytical needs.”
The modeling capabilities and calculation engine far exceed what Mischler was able to do with Excel: the firm can separate data from the logic of the presentation, which means it can house data in various places without dragging down performance. “I have data stores that are two gigabytes in size – imagine trying to put them into a spreadsheet and get some intelligence out of them,” says Klinghoffer. “The memory requirements alone would shut them down.”
In addition to improved delivery and calculation, Mischler says plain language formulas and efficient structure makes it much easier to manage models. “Even with such large data sets and complex models, I can come back in after months and understand what’s going on,” says Klinghoffer. “I can step through a formula to see what is being manipulated, and understand the impact of any change I will make. Thanks to category relationships between matrices and dimensional linking through formulas, even my most complicated models only need a few hundred formulas to calculate tens of millions of cells!”
In the spring of 2016, Mischler rolled out its package of models to customers via the Quantrix Qloud, available 24/7, including portfolio and security analytics, investment strategy development, and inventory offerings. It offers these tools for free to clients who execute some of their bond transactions with Mischler, providing intelligence and data that were formerly out of reach to them and leveling the playing field for Mischler with the largest investment firms.
“Clients can buy their bonds from anyone – this is our value-add,” says Klinghoffer. “We give them the tools they need to make decisions, all within a seamless online cloud environment. This is a slam-dunk for us – and we’re just getting started.”
Mike Salisbury is the Director of Quantrix and is responsible for the overall Product Roadmap for the Quantrix Modeler and Qloud products. Mike took a few minutes to share some insights on recent product developments and upcoming new features.
It has been a busy few months for the product teams. Can you give us an update on recent product developments on the Desktop modeler?
Certainly! In January we released a major version of the desktop product and I am pleased to say it is getting great reviews from our user community. We totally reskinned the User Interface application and focused on making the commonly used features much easier to use. New icons, new color themes, new launch panel, all aimed at making the most seasoned user to the new user more comfortable with our application.
Our Dependency Inspector (DI) redesign is a big hit. It makes formula debugging much simpler and users of all abilities can trace the calculation thread throughout the model. I personally use the Salesforce integration daily in our Quantrix sales forecasting model, and our Tableau integration means users can get information directly out of Quantrix into Tableau reports.
In our March release, we enhanced the DI even further by allowing users to navigate directly to the formula from the DI to make formula edits. Users can re-index and rename their category items on-demand, and we continue to make strategic UI improvements like shading the formula edit area in a light green to make it clearer to the user where the focus is on the screen.
How about recent Qloud developments?
We are really excited about our IOS tablet integration. I encourage iPad users to navigate to http://qloud.quantrix.com to check it out. Having the ability to use Quantrix Models on a Tablet device should open up some opportunities for executive and ‘in the field’ type use-cases we have in our customer community. I think one of the best features is being able to move a matrix dimension tile to the row, filter, or column tray simply by dragging it with my finger. Simple but very effective!
In our March release we introduced Salesforce compatibility for Qloud. If you are using the Salesforce integration on your desktop models, you can now use this integration on both Hosted and Enterprise Qloud. We believe the data analysis you can do in Quantrix far exceeds anything you can do within the Salesforce application itself.
How often do you release new software?
We strive to have a new desktop and Qloud release every three months. We have adopted the Agile method of software development. This allows us to quickly iterate on new features (and bug fixes) in two-week cycles or ‘sprints’. At the end of each sprint, we have a functioning product. At the end of the three-month cycle we move into a formal test period and then release to our customers. Doing this allows us to be much quicker and more reactive to our customer needs.
What is next for the Desktop Modeler and Qloud?
In regards for Qloud features, we are actively working on Multi-user Edit. This will be an Enterprise-only feature that will allow Enterprise Customers to deploy models to users, and then allow those users to edit the model at the same time – no more waiting for your colleagues to release the lock on a model. With this comes some powerful new features including tagged versions and ‘save-as-you-go’, all to ensure you can collaborate effectively.
For Desktop, continuing the improvement of our user interface and ease of use as well as developing more getting-started help is high on our list. We are making our Salesforce integration even better by allowing joins between multiple tables in the Salesforce hierarchy. We continue to address issues and enhancements to make sure our calculation engine is as efficient as possible, because we know our customers like to make very large models!
How can a Quantrix user provide feedback?
We very much appreciate feedback from our users. Many of the new enhancements and features are the direct result of user suggestions. There are many ways to provide feedback:
- For our customers with a support contract, you can email firstname.lastname@example.org. Feedback and suggestions are logged in our support systems and you are notified when they are addressed and released.
- Email email@example.com with any suggestions.
- Post in our on-line forums at quantrix.com
- Contact me (firstname.lastname@example.org, +1.207.420.3101) or Product Manager James Kipling (email@example.com, +44 1483 595 000)
- Attend our annual Seminar by the Sea user conference!
As always, thank you everyone for your support of Quantrix!
At last fall’s Seminar by the Sea conference for Quantrix users, Rich Holohan of LogCapStrat sat down with us to talk about his use of Quantrix for quantitative modeling. Like many former Excel users, he needed a replacement for an Excel model that had “gotten out of control.” In this video, he talks about the flexibility and efficiency of Quantrix, including the dramatic reduction of man hours required to update one client’s model. From three people full time for two weeks, to one person for two days, the Quantrix model offered “massive cost efficiency.” Watch the full video for more.
Brian is the Engineering Manager and Lead Engineer for the Quantrix team and has been with the company for over ten years. He has actively participated in the design, production and testing phases of each of the releases of Quantrix since version 1.0. He currently manages the engineering team and is the lead engineer on the various projects for Quantrix Modeler and the Qloud.
What is involved in your role as lead developer for the Quantrix Modeler desktop product, and how has it changed over the years?
As Engineering Manager, I help manage the prioritization of issues and run the projects from an engineering standpoint, as well as other typical managerial duties. I also do a lot of coding of features, which I enjoy. I’ve been with Quantrix since 2003 and from an engineering perspective, a lot has changed. Java, the base language we code in, has improved. User interface design principles have changed, particularly in terms of migration to mobile devices. In that period, we’ve moved from being a desktop only application to the web, launching the Qloud, and continuing to move to mobile device. And we have had a lot of fun.
What is most important focus from an engineering standpoint?
Two things: usability and stability.
Talk about usability.
In the 1990s, usability meant, “Can you do it?” Today, it’s “How easy is it to do it?” That is a big shift. We want to make new users more comfortable using Quantrix. Power users love it, but we are working on making it more usable for more people. That means making things easy to find and easy to do.
Some recent examples of our work on usability include the recently improved launch panel and the major changes to the Dependency Inspector. Before, it was hard to understand what was happening in the cell. Now there are a lot of clues as to how it calculates, including the ability to hover over the formula in the dependency inspector to see the calculations. Another change is that when a user starts to edit a formula, it gets a green background. You know exactly what you’re editing. We plan to do more usability improvements like this as well as revamping some other tools.
How important is user feedback?
We have a great user base, and they provide great feedback. We also track when people call support. If they have a hard time doing things, we try to fix that in development. We try to be as responsive as possible. At one of our Seminar by the Sea events, a power user told us that it would be great to be able to select a range of cells and filter on them immediately. One of our programmers went home that night and coded up a proof of concept. We included that feature in our next release.
Have you always been a software engineer?
I served in the Navy on submarines. Then I became a high school math and science teacher. I went for a masters in computer science, and learned about Quantrix. I pestered the company founder until he gave me a job.
What do you do when you’re not coding for Quantrix?
I have a lovely wife and two young children. We love to go cross-country skiing and snowshoeing. I also love to read – I’m currently reading a fascinating history of Maine called The Lobster Coast.
Dan Kirkpatrick of Validus Holdings Group attended the 2015 Seminar by the Sea in Portland this fall, and shared his thoughts about Quantrix Modeler. In this video, he discusses how the company uses Quantrix for financial planning, including developing P&Ls and balance sheets by company and division. It also uses Quantrix for scenario testing and for building capital models. Validus is a holding company for reinsurance and insurance operating companies and investment advisors, and it has a high volume of historical data – its models are large, with some at 1.2+ billion cells. Kirkpatrick says that Quantrix is “leaps and bounds ahead of Excel” in its ability to create models that deliver insight to senior management.
Nathan Briggs, senior treasury analyst at Jackson Laboratory, uses the Quantrix business planning and modeling solution to manage endowment and treasury portfolios. The models track investments, accounting information, and market variances along with return calculations.
Jackson Labs switched its models from Excel to Quantrix to provide a more robust and structured solution, and improve formula management. “What we liked about Quantrix was the ability to change one formula in one place and have it be pushed throughout the model,” says Briggs.
In addition to desktop modeling, Jackson Laboratory uses the online version – the Quantrix Qloud – to improve collaboration with external investment advisors and managers. “Using the Qloud has allowed us to share information more efficiently,” says Briggs. “Now we’re simply publishing that information to the Qloud – they can go on when they are ready and download information and manipulate it how they need to. It certainly has been a time saver.”
Dan Murray is Director of Strategic Innovation for Interworks, a global IT services and business solutions provider. He is the author of Tableau Your Data!, and will be the featured speaker at this year’s Seminar by the Sea. We had a chance to talk with him about the data modeling ecosystem, and the value of an iterative process.
Describe your practice at Interworks.
Murray: InterWorks has a different model than most consulting companies —we aim to do high quality work, very quickly and then leave. Our consultants are good at translating technical needs into business cases. We also provide clients with thought leadership and project planning/implementation. Then we help them design dashboards that help them see and understand their data. We
have a broad skillset surrounding business information including data visualization, database design, extract, transform and load logic, hardware and network design to enable appropriately responsive data.
How did you get started working with data?
My first job was with a leveraged buyout firm. We bought bankrupt companies, looked at data, figured out how to make them profitable, and flipped them. That was better than an MBA – I figured out how to set up capital structures, leverage business assets, and drive profits. Over the last 30 years I’ve used data to solve problems, identify opportunities, and drive improvement. In the last seven years I’ve focused on teaching clients how to do this well.
What are some of the changes you’ve seen in your 30+ years managing data?
The proliferation of data bases – they are growing all the time. Many of our clients don’t have one database; they have a dozen and they can’t figure out how to get to the information. Our practice is built on helping companies bridge those gaps. And the tools are so much less expensive. Small and medium-sized businesses can’t afford to spend $1 million on toolsets to extract data and spit out multiple scenarios. Tools like Quantrix are dramatically lower in cost.
What else sets Quantrix apart?
Quantrix defines formulas the way that databases think – not by the cell, but by dimension. You can take a model with 4,000 formulas in Excel and do it with 40 in Quantrix.
You talk about the data ecosystem – where does Quantrix fit in?
At the center of everything is some kind of database. And Quantrix does something that no other tool does – it architects formulas in a very efficient and powerful way by speaking the database language at the data level, not the cell level. This means it can do powerful things with assumptions for predictive analysis. Factors that affected a business’s performance a year ago might not affect it a year from now, so you need to build a complex calculation that based on a sliding assumption. You identify the dimensions that are important to your business, and the assumptions that have worked in the past. With Quantrix, you can adjust the sliders to see if each factor will impact the future, by how much, and the outcome.
Give me an example of how that information impacts performance over time.
Today, most companies create a budget, and then re-budget every quarter. They take it for granted that they need to revise their forecast, but they are not using that information to refine their forecast. Quantrix can be the tool that you are iterating the potential future from. When you have new information, you should be populating the database with that innovation, and capture your assumptions. They are part of the innovation cycle. We tend to forget the bad ones and overemphasize the good ones – but with Quantrix connected to a database, you can see how good your assumptions were, and use them when you iterate. If we can reduce the time and effort required rebuild that forecast, then you can really advance over time.
What’s the bottom line for companies?
In the end, it’s all about data. Do you have it? Can you interpret it correctly? If you move the dial – which you can do easily with a Quantrix slider – did the expected outcome occur? That, to me, is what it’s all about. Drop the technology hurdle, expose the bones from the assumptions, and have a learning process where you get better over time. It is measurable. At some level, something significant happens – if you’re doing it right, the cumulative effect of the effort will help you grow.