During the Banking Royal Commission, I participated in an ANU Committee that investigated the viability and role of aggregation in the mortgage market, with findings fed to the commission in order to shape their discussions. Our recommendations weren't published, discussed, or considered. Instead, it was the focus of broker commissions, best interest duty, and the need for trail commissions that dominated commission conversations and media commentary. What the investigation did prove is that the existing aggregation model is vulnerable to massive disruption without a significant operational impact on brokers. In this article we'll speculate on what some of these changes might or should look like and consider the possible impact on the mortgage industry.
The mortgage industry is very much a dinosaur. The aggregation model, in particular, is a metaphorical steam-driven swiss-cheese ridden Titanic driving at full-speed towards an iceberg. The difference for us is that the iceberg is in full view. What are we doing about it? We may not know precisely what any change will look like, but we can bet with certainty that the next few years will introduce changes whether we like it or not.
This article provides no answers or solutions. One of our core company objectives is to have aggregation groups reshape their model into one that would provide continued value without a core aggregation product. If aggregation as we know it wasn't necessary tomorrow, what what these groups become?
What is an Aggregator: By definition, an aggregation is a group of brokers, and "aggregation" is the process of negotiating and resolving supply and pricing for those in the aggregated group. Funds are aggregated and then paid out to brokers. All that is required to set up an aggregator is an ACL and a number of downline brokers that the lenders are willing to negotiate. The process of setting up an aggregator is not necessarily difficult, and it should often be considered in company with franchising.
Making the claim that aggregation isn't needed is a big one, but it isn't without precedent. Open Banking was a first and essential step in establishing that all ADIs had the capacity to conform to industry standards, and in response to Consumer Data Rights 
legislation they've demonstrated that they're able to implement an open-sourced standard in a cooperative manner. It could be argued that the same cooperation could be applied to scale the Open architecture to cater for broker connections and commissions.
A style of 'aggregations' will always exist. While payouts could and probably should be made to ACL holders directly (it's no longer 1987), there will always be representatives that will lean on a parent group for managed commissions. For Credit Licence holders, the monthly fee or split that is currently paid is one expense that might be eliminated via an alternate model, and this would open up more freedoms and flexibility with regard to trail 'ownership' (the ownership of trail is somewhat of a misnomer given payouts are made to an entity other than the intended end recipient). Regardless of what you're told, you don't "own" your trail.
The loan application itself currently feeds data into a proprietary lender endpoint (or in many cases simply reverts to an email) via middleware, and it's here where the mortgage industry has most amount of room to move in order to significantly reduce costs, thus providing a better deal to the end consumer. Building these endpoints into the Open Baking architecture is a logical next step, and it removes control from a virtual monopoly that's under no pressure to improve. The fact we don't have a shared loan application architecture is an embarrassment to the industry that needs to be resolved sooner than later, particularly when broker channels effectively own the residential market.
With increasing pressure on aggregators from lenders to ensure compliance within their ranks, it's the compliance obligations that tends to dominate their current responsibilities - even if it's not as visible as the 'extra' services on the periphery. However, the bread-and-butter mandates aren't something that has to be difficult, and oversight by real-time AI has the potential to mitigate and/or identify infractions - intentional or otherwise - that would otherwise go unnoticed, and in company with real-time human auditing supported by more modern software methods we can identify virtually all categories of malfeasance with higher-risk files flagged for evaluation.
Compliance: Compliance auditing can be outsourced for a very low fee; we don't necessarily need aggregation to provide this services (which some of them do very poorly), and lenders would invariably prefer the former in order to take advantage of the expertise. Aggregators can't lean on their 'compliance' speciality when we're already seeing AI that does a better job.
In order to prove a point, we built a full-featured RESTful API that'd be used by various stakeholders to submit and manage applications of all types. Built on top of the existing Open Data framework, and by using existing methodology, we built a proof-of-concept product that demonstrates how millions might be slashed by adopting a shared open architecture. If a similar system was managed with an enthusiasm by lenders, something similar might also manage broker commissions, trail, splits, accreditations, clawbacks, and support, thus reducing the man-hours associated with caveman human handling. It's obviously not up to us to manage industry tech in this respect, but we'd love to see our industry bodies act in the best interests of brokers by working towards an industry-owned non-profit 'clearing-house' style of solution that would ultimately give brokers longevity with forwarding-thinking technologies.
Lender Accreditations: Accreditations with lenders will rarely vary between groups, yet migrating an accreditation from one group to another can often take weeks. In some cases we still see printable PDF documents used as part of the process - a caveman solution that is in opposition to the digital frameworks we expect in a quality-controlled environment. Again, accreditations should be assigned to a broker as part of a shared open standard, and those accreditations should be easily migrated between licence holders.
Online Applications: We've taken our application model and modified it so it's suitable for a simple website feature that supports online applications. Our clients will see a watered-down version of this API made available sometime in the future.
We're kidding ourselves if we believe our own thinktank-style solution will have any traction. When I used the iceberg metaphor with a group of industry leaders over lunch, one aggregator CEO nodded his head, agreed that we're sitting on a virtual ticking time-bomb, and said that the best we can do is pick the smallest and prettiest iceberg in the ocean. Another top-5 CEO suggested that we start investing in bigger buckets. Given that big tech took over the music industry virtually overnight, and given what we've seen with streaming services taking control over video distribution, shouldn't we start to identify the bigger-picture risks associated with medium-term industry survival?
At this point we haven't mentioned the single issue that potentially overshadows all other threats to our comfortable dinosaur model: AI.
Some time back we conducted an early experiment in AI capabilities 
by replacing a broker with artificial everything 
. We didn't have the resources to scale our concept into something bigger, but we certainly expected somebody else to do so. Fast forward to now and we're still not seeing much from AI beyond the standard vanilla 'let is summarise your notes' style of thinking. That said, AI represents the single biggest threat to our industry.
We often hear that "borrowers will always want to talk to a human", and while there's elements of truth in this statement, we have to remember that the guys at Blockbuster bet their entire future on the belief that we'd want to browse shelves, and the guys at Brashs Record Stores probably thought that we'd always want to own a CD. Then there's the guy that bought his taxi plates the week before Uber hit the streets, and there's a poor sap at Kodak that bet the company's future on film. In all these cases we couldn't reason with the opinions of the decision makers because they knew they were right, just like we knew the Earth was flat 500 years ago. Our reality is seriously distorted by a perception, and we can't let personal bias or experience blind us from reality or impair our ability to think critically and rationally. It could be argued that the whole "they'll always want a real person" thing is merely a self-serving delusion. What evidence actually supports the claim? Make no mistake: AI is faster, smarter, provides better customer outcomes, and it provides a far better post-settlement and repricing service. Ignoring the impending threat is ridiculous. Instead, we have to make AI our slave before it makes us theirs.
Technology takes no prisoners, and it almost always wins. What are you doing, and what is your company doing, to stay ahead of the impending advance? More importantly, what is your aggregator doing to ensure that you have all the tools you need in-house to satisfy your existing and emerging AI requirements? What is the industry doing to ensure it has a defence against the advance of big tech?
We're in a unique position in that we're connected to the technologists that want to disrupt our existing retail aggregation model so we're acutely aware of their impending plans to decimate our current remuneration pathways and architecture. As an industry we've tightly clung to a model that has stood still in time despite the emergence of technologies that could replace the entire infrastructure overnight. Instead of allowing our modelling to be controlled by big banks and tech, we should be putting pressure on our industry bodies and aggregation groups to develop a shared, common and fully-owned framework that will support the industry into the future.
If we want aggregation to survive, we have to start building our new future today.
on Google Maps). Shown here as the 'Mulberry Tree' Bed & Breakfast (source: Google), still branding the 'Bank of Australasia' markings. [ View Image ]
