Imagine buying a slice of a pool of loans. The borrowers send in their payments, and the money reaches you. A pleasing arrangement, until you ask the awkward question: who else is standing in line?
Some investors may be paid before you. Some payments may be diverted when a test fails. Borrowers may refinance, repay early or stop paying altogether. The security’s documentation sets the rules. Intex Solutions turns those rules into cashflow models, allowing financial professionals to examine what a particular slice of the deal would receive under different assumptions. A legal document has become a machine you can interrogate.
- The work: model the payment rules of structured securities.
- The buyers: banks, investment managers, issuers and other institutions.
- The advantage: reusable deal models, maintained data and multiple ways to access them.
- The catch: an exact calculation still needs sensible assumptions.
That is an unusually specific business. It is also a useful way to understand a corner of finance whose vocabulary seems designed to discourage visitors. Before worrying about the alphabet soup, follow the money.
A queue disguised as a bond
Structured finance gathers assets, such as mortgages or other loans, and creates securities with different claims on the resulting payments. Those classes of securities are called tranches. Their payment priorities form a waterfall. The waterfall can be complicated, but its central question is wonderfully ordinary: who gets served first?
Intex’s core territory includes residential mortgage-backed securities, or RMBS; commercial mortgage-backed securities, or CMBS; asset-backed securities, or ABS; and collateralized loan obligations, or CLOs. Each abbreviation points to a different kind of underlying exposure. Each deal brings its own instructions for distributing cash.
Intex models transactions individually using the amortization and loss-allocation rules in their documents. An analyst supplies a scenario, and the engine calculates the consequences. Faster prepayments might shorten the period during which interest arrives. Defaults and recoveries can change the amount available. The investor needs to examine their particular tranche, rather than assume that knowing the pool means knowing the bond.
Same pool. Different places in line.
The spreadsheet grew a memory
Intex started in 1985 with spreadsheet add-ons for Lotus 1-2-3 and, later, Excel. It introduced a bond calculator in 1987 and mortgage-backed securities calculation tools in 1988. By 1990, it had developed its proprietary CDI deal-modeling language and released initial agency deal libraries.
The progression matters. A calculator helps you do a calculation. A library saves you from reconstructing the same transaction whenever you want to ask another question. By 1992, Intex had expanded into non-agency deals and enhanced the language to handle their default and loss features. The fine print was acquiring a reusable form.
Carnegie Mellon identifies Mark Gelfand as a founder and records his physics degree, earned in 1973. That technical background is an agreeable detail in a company whose product is so much about translating one kind of precision into another.
Intex introduced Windows software, INTEXdesktop, in 1997 and the web application INTEXnet in 1998. DealMaker followed in 2000. The company’s history records its first European transaction in 2003 and a London office in 2006. The interfaces and geography changed; the need to express the rules of a deal remained.
The model needs a second job
On its current website, Intex says it has independently created and actively maintained models for more than 50,000 deals. It also reports close to complete coverage of US deals in supported sectors and structured securities issued in the UK and continental Europe. These are company-reported coverage claims, rather than a promise that every imaginable private transaction is available.
deals modeled and maintained
The less glamorous verb is the important one. A deal keeps changing after its model is written. Intex’s data architecture distinguishes CDI files, which contain relatively static payment rules, from CDU files, which carry regularly updated deal information. A useful model needs both its instructions and its latest condition.
The hiring pages make that work unusually visible. An ABS updating analyst helps own the monthly update process for thousands of transactions, calibrates datasets, works with data providers and supports automation and quality assurance. The modeling role constructs and calibrates cashflow models and helps adapt the modeling language to new structures. This is software with an ongoing editorial obligation to the facts.
Client support is technical work too. Intex’s client-services role includes testing new features, maintaining help material and conducting training. Its careers page describes casual dress and a relaxed culture. The job descriptions supply the more revealing detail: attention to accuracy is a professional qualification.
One engine, several doors
INTEXcalc gives an end user a place to analyze individual securities, bid lists and portfolios. DealMaker serves the earlier moment when a transaction is being structured: users define assets and payment waterfalls, then can distribute a proposed model for investors to test before pricing. APIs put the cashflow engine into custom systems. INTEXlink brings data into Excel.
The formal INTEXcalc launch announcement in February 2011 included a testimonial from Cantor Fitzgerald director Phillip Daskevich. The attraction he described was practical:
“I was up and running with INTEXcalc for the first time in just a few minutes”
Phillip Daskevich · Cantor Fitzgerald · 2011 launch announcement
Intex also works inside other companies’ applications. Its current integration directory lists BlackRock’s Aladdin, FactSet, Murex, Numerix PolyPaths and other financial systems. The commercial logic is straightforward: an institution may want specialist cashflow analysis inside the system where it already manages portfolios or risk.
A 2015 agreement connected Misys FusionInvest with Intex’s models and data. In 2016, dv01 announced an integration that let mutual clients import marketplace-loan pools into INTEXcalc’s Loan Portfolio Tool. Those are historical examples of the same useful habit: meet the analyst in the workflow.
An expensive subscription to fewer surprises
Intex sells to institutions through subscriptions and software and data licensing. One concrete glimpse of the economics comes from CalSTRS, California’s teachers’ retirement system. Its quarterly contract report lists an Intex subscription renewal dated September 1, 2020, for $167,739.
CalSTRS subscription renewal, 2020.
A disclosed contract amount, not a current price quote.
For a buyer, the question is what work the subscription replaces or supports: constructing deal models, keeping data current, repeating scenarios and fitting those outputs into institutional systems. The subscription does not purchase good investment judgment. It purchases tools and information with which to exercise it.
Alternatives depend on the task. Bloomberg and Moody’s Analytics appear alongside Intex as information services in a 2024 CLO indenture. CMBS transaction materials also identify Trepp as a modeling provider. An institution can build its own models. Broader analytical platforms may integrate Intex as well, making the market less tidy than a simple list of rivals would suggest. The relevant comparison is the work covered by each product, the deals available and the effort required to use it.
The assumption still belongs to you
Here is a miniature version of the problem. Suppose a hypothetical deal has 100 units of cash to distribute, with a senior claim on the first 70 and a junior claim on the next 20. Reduce the available cash and the investors do not all lose the same amount. Move the slider to see the order matter.
How much cash reaches each claim?
In practice, analysts can change assumptions about rates, prepayments, defaults and recoveries, then examine projected cashflows and valuation measures. A sensible workflow compares scenarios, inspects the assumptions producing an attractive result and checks the freshness of the data. If a projection depends on borrowers behaving implausibly well, more decimal places will not rescue it.
The company’s public product material lets users control assumptions and incorporate third-party prepayment and credit models. That leaves a clear boundary: calculating contractual consequences is different from deciding which future to expect. Stale data, a missing deal model or a poorly chosen scenario can undermine the exercise before an investor ever makes a trade.
For readers outside finance, there is a business lesson here worth stealing. Make difficult expertise reusable. Keep the information current. Give customers several ways to reach the same engine. Those principles are an inference from Intex’s products and history, and their usefulness depends on a market where the underlying rules matter enough for people to pay for careful translation.
Intex’s specialty is to make that translation available for inspection. Someone still has to choose the assumptions. At least they can see where the money goes.