FIX Software: Past, Present & Future

Chris Lees 27th Mar 2019
5 min read

API FIX Onboarding

One of the most important and difficult tasks for any software organisation is to find what is known as "Product/Market fit".

“Product/market fit means being in a good market with a product that can satisfy that market.” Marc Andreessen

This simple idea contains two simple ingredients, a "good" market and a product, both of which most companies will suggest that they have. And yet so many firms struggle because they haven't found their correct product/market fit yet; a great product in a bad market will not be a success, nor will an incomplete or bad product in a good market.

We are all outside observers on lots of markets, affording each of us the luxurious freedom of being able to form (largely uneducated) opinions on the overall shape and direction of various markets or products. Whether our opinions, conclusions and ideas are accurate or achievable or not is irrelevant because - thankfully - none of us will ever really have to action any of them!

When you work inside an industry, however, practical realities set in and it becomes much, much more difficult to determine the root cause of inefficiency, gauge overall industry direction, and to spot opportunities. Some people say it is difficult to see the wood for the trees, but I think it is mostly down to the higher level of understanding that those inside an industry have compared to outsiders - we know what currently works and the constraints applied upon us. Such industries have so many unknown or conflated issues, however, that building successful products is tough and finding that perfect product/market fit is often elusive.

This is the reason why - from the outside - the business of financial services connectivity can seem antiquated and inefficient, and simplistic uninformed solutions such as "let's just replace FIX with modern web APIs" are unlikely to gain traction.

For the past year or so I've found myself returning to this idea of product/market fit time and again, as a nagging voice in the back of my head keeps telling me that something isn't quite right... something isn't clicking the way it should.

The same, unanswered questions and observations keep circling around in my head:

  1. If everybody really does "speak" FIX, then why does it continue to be so hard, manual and error-prone to make connections, and why are they held together with sticky-tape?
  2. The "market" for FIX should be expanding as new venues arrive faster than others close, but it often doesn't feel like this. After 20 years of experience, why haven't onboarding times shrunk, and why can nobody ever seem to give an accurate estimate of how long a new connection will take?
  3. On one hand, new venues talk about the importance of building a client base via a select group high-profile vendors, but then loudly complain about the very long lead-times, limited functionality and high cost. Why have we all become so reliant on gate-keeper vendors with so much power they can determine the success or failure of start-ups?
  4. Firms routinely complain about hair-raising, monopolistic price hikes from legacy vendors seeking to exploit what they (correctly) consider to be extremely sticky products. In any other industry, the natural response would be to seek an alternative product... but without new entrants, there is nothing to replace them with and therefore the only option would be an unpalatable in-house build. Why is the market so unattractive to new entrants?
  5. Some would suggest that the future is not do-it-yourself connectivity, but instead "managed services" (often those people work for managed service providers!). Aside from the fact that this compounds issues 3 and 4 above, their potential customers also suggest that the attraction is about the "pull" of new service benefits, but rather a "push" due to fatigue of their existing processes.

So does this add up to a "good" market in Marc Andreessen's definition?

Well, the FIX market should be big - thousands of firms worldwide - and the mechanics of building and maintaining connections are sufficiently consistent that well-designed products should satisfy the majority of participants. It's also fair to say that there is a great deal of inefficiency and fatigue in areas that are facing continued budget pressure and management scrutiny, which should create a demand for efficiency-boosting solutions.

So the FIX market should be a good one even though it may not always feel like it! So what would a good product/market fit look like here?

To answer this, I believe we should take a moment to reflect on the fact that in 25 years of FIX and electronic trading there really has only been two significant commercial vendor products directly related to FIX connectivity - the CameronFIX engine, and VeriFIX for testing. (Before anybody comments that there were other notable products such as UL Bridge, NYFIX network, algo tools or OMS/EMS's, I'm not disputing that other products may have been profitable in the FIX space - my point is that there are really only two widely-adopted, widely-discussed, stand-out products that captured very high market share in their heyday).

What links them? A FIX engine and QA tools are both mandatory for operating a FIX API, and they were the first to satisfy the market for these particular problems during the early boom years of FIX adoption. They had a good product/market fit and great market timing... but mostly they were necessities.

My hypothesis is that arrival of CameronFIX and VeriFIX early on in the market satisfied two immediate needs for mandatory components, and firms simply added headcount and free tools to deal with everything else - writing documentation in Word, crafting XML files in open-source editors, and grepping logs by hand. And that is pretty much where the market has remained stuck since the 2000s. New vendors find it difficult to displace such deeply-embedded products and the processes that surround them, and there is insufficient free "market space" in this limited function to build additional, value-added tools that can't be handled - albeit very inefficiently - with free alternatives such as Word or Excel.

Are there any external factors changing the industry now?

  • Firms certainly want to embrace the cloud, but that's a cost play and doesn't really make them more efficient.
  • Firms continue to move away from high-cost legacy products towards lower-cost alternatives, again cutting costs but also not really making them more efficient.
  • Firms can launch programs to try to remove some historical technical complexity, but this takes resources and the benefits are not felt immediately.
  • Modern specification editors and tools to automate certification exist, but there is an initial resource investment required and many firms were burnt by poor-quality legacy products.
  • FIX Trading Community is pushing machine-readable specifications but as I've argued elsewhere this is not a silver bullet, and firms that I talk to do not believe that they will be early-adoptors here.

So where do we think the market is headed and how can we help?

I truly believe that the key to unlocking a new wave of efficiency in FIX does not lie in replicating existing software or some magical machine-readable format. Instead, I believe it will arise from redefining the boundaries of what it means to be in a vendor in the FIX/connectivity space and the services they offer. In particular, I believe that it will come about with new tools and processes which assist firms of all sizes to build and operate efficient connectivity workflows.

Explicitly, we believe the future lies in:

  1. Helping firms to develop agile, repeatable, efficient and predictable connectivity processes
  2. Helping management at those firms to regularly review progress, maximize resource usage and iteratively improve
  3. Processes and solutions which extend beyond just considering FIX, into wider connectivity tasks such as organising telecoms lines
  4. Processes and solutions which handle not just initial customer onboarding, but also organise production issues and projects

If you have ever come across HubSpot you will recognise that they invest a lot of time educating customers about how best to use their software within a wider inbound-marketing workflow. Similarly, Kanban is a powerful "agile" planning strategy on it's own, but it is the combination with the JIRA software that makes your development cycles more efficient and predictable. Or what about Salesforce, which moves from shared database to an efficiency-boosting platform when surrounded by well-designed, efficient and repeatable business processes?

Software and process, hand in hand. This is where FixSpec is headed.

We are pivoting towards helping firms to define and operate efficient connectivity processes and teams. Our software tools such as SpecServer, Central and SideKick all form part of that endeavour, but the ambition is wider. So watch out for us starting to talk more about process and best practise in the months to come, and stay tuned for practical help and advice on how to improve your connectivity process.

The future is bright!

Find This Useful?

To receive more tips for improving efficiency in YOUR connectivity process, sign up to our FREE monthly email newsletter.

Awesome - thanks for signing up. You can unsubscribe at any time by simply dropping us a line.