Secondary Marketing & Capital Markets Software

Execution is the whole purchase

Buy the system that moves your execution, and treat everything else as a tie-breaker. One basis point on a billion dollars of annual volume is a hundred thousand dollars. No contract in this category comes close to that number. That is why we weight production impact and depth of function above everything else here. A hedging system that is pleasant and wrong destroys far more value than it costs.

The rest of the scoring follows from that. Integrations matter because a missing agency connection becomes a manual workaround on your worst day. Support matters less than in other categories. A secondary desk will tolerate an ugly screen. It will not tolerate a position report that is wrong at nine in the morning.

Best efforts to mandatory is the first decision

Most shops looking at this software are really asking whether to go mandatory. The software question follows the execution question, not the other way around. MCT publishes the historical figures it sees on that shift. It cites a 20 basis point pickup on conventional volume and 40 basis points on government volume. Those are MCT's own numbers, and they set the bar every competitor gets measured against.

The cost side is public too. Fannie Mae's whole loan platform charges extension fees and pair-off fees on mandatory commitments. It also applies a duplicate price adjustment when you recommit a best efforts loan inside 30 days. Fannie states plainly that fallout on cancelled or expired best efforts commitments affects your future pricing. So the choice is not free money. You are trading a known price concession for unknown fallout risk.

Ask each vendor to model that trade on your own pull-through history, not on a demo file. Ask what happens when a big loan falls out three days before settlement. The answer should be a specific workflow, not a phone number.

What the hedge system has to do on a bad morning

Coverage mechanics are where these products separate. You need a position that reflects real pull-through, not lock count. You need the mark to move with the market before you set tomorrow's rate sheet. You need pay-ups and roll cost carried into the profitability number, not bolted on later.

Optimal Blue publishes results from its own impact study on CompassEdge. It claims more than $1,000 average net benefit per loan. More than $400 of that comes from hedging and trading. It also claims a 43 percent increase in operational capacity. Treat those as vendor figures and ask how the study defined the baseline.

MCT has pushed further into automated recommendations. Its Atlas advisor now produces hedge recommendations against live pipelines. Optimal Blue has answered with a set of embedded assistants for position, profitability, projections, and trade ideas. Both firms say these carry no separate charge. Both are new enough that you should test them against a week you already lived through.

Whatever the model suggests, the desk still owns the trade. Ask whether the system logs the recommendation and the human override side by side. That record is what you show a board when a month goes badly.

Settlement discipline is a feature, not an afterthought

TBA hedging runs on a calendar that does not care about your pipeline. SIFMA sets four monthly settlement classes, A through D, each with its own notification date. In January 2026 the Class A notification fell on the twelfth and Class D on the twenty-third. Sellers must notify buyers of pool information at least 48 hours before delivery. Miss the cutoff and delivery slips two business days.

Software that treats notification as a spreadsheet task will cost you eventually. Ask where the trade blotter lives. Ask whether the system knows the class calendar or expects your analyst to know it. Agile Trading Technologies built its business on this layer. Its platform covers electronic TBA request-for-quote, pool bidding auctions, limit orders, and dealer-to-dealer trade assignment. Agile publishes very little in the way of numbers, so verify counterparty coverage against your own dealer list.

The bid tape is where the basis points hide

Hedging protects the margin you locked. The loan sale is where you win or lose the rest. Bid tape assignment of trade is the execution most worth understanding, because it removes the pair-off from the cash sale. MCT reports average client savings of $97,000 per year in 2023 on that execution. It also holds a patent on committing to buyers as a spread to the security, not a fixed price.

The tri-party agreement between the seller, the buyer and the dealer used to be a fax exercise. Agile automates it inside its trade blotter. MCT automates it inside Marketplace. If you sell any real mandatory volume, this single workflow deserves its own demo session.

Polly comes at the same problem from the trading side. Its Loan Trading Exchange offers a code-free eligibility editor. That builds one-click bid tapes, with live TBA benchmarking on returned bids. It connects by API to Fannie Mae, Freddie Mac, PennyMac, and other buyers. Polly describes itself as hedge-agnostic, and its exchange pages do not offer a pipeline hedging module. That is an honest position, and it explains why Polly sits lower in this category than in pricing. If you want one vendor covering the lock through the trade, Polly expects you to bring a hedge advisor.

Agency plumbing ages faster than you expect

Agency commitment APIs change several times a year, and every change is a delivery deadline. The vendor's shipping speed is a real buying criterion. MCT gives a usable measure of its own pace. Freddie Mac announced an LTV specified pool pay-up category on 27 February 2026, and MCT deployed it by 10 March. Freddie announced a manufactured housing pay-up category on 23 March, deployed by 30 March. MCT also added support for Freddie's Co-Issue All-In Funding API for loan-level pricing and committing, effective 30 March 2026.

Optimal Blue is building the same muscle from a different direction. Its Agency Direct release targets retained execution to the two agencies and the Federal Home Loan Banks. The pitch is central loan import and commitment management in place of spreadsheet tracking. Coverage modeling sits in the same screen. Several 2026 Summit items are live and others are still in phased rollout, so confirm what you are actually buying.

One more detail worth knowing before you negotiate. Fannie caps combined daily committing at $200 million before you have to call the desk. Its extended best efforts window runs to 11:30 pm Eastern. A system that cannot commit inside those windows without a human is a system that will cost you nights.

Warehouse cost is the trade you forgot to price

OptiFunder is the odd product in this ranking, and it earns its position honestly. It is not a hedging system. It sits between funding and loan sale, where interest expense quietly eats the gain you fought for. Genesis serves originators and Greyhound serves warehouse lenders. The company reports 100,000 loans funded monthly and around 5,000 automated touchpoints a day. It holds real-time connections to more than 60 warehouse lenders.

Its allocation engine picks the line for each loan using capacity, sub-limits, product eligibility, rates, non-use fees, and rebate incentives. OptiFunder claims an average 10 percent reduction in warehouse expense for independent mortgage banks. Test that claim against your own last quarter of funding data before you sign anything.

The workflow benefit is easier to verify than the savings claim. Batch wire requests and purchase advice reconciliation are measurable in staff hours. So is collateral tracking. One published client account describes 30 percent more volume on the same four-person funding team. If your funders are working late every month end, this is the cheapest fix on the list.

MERS is not really on your shortlist

MERS scores well because it works, not because it competes. It is infrastructure, and you join it rather than choose it. The eRegistry is the legal system of record identifying the controller and the location of an eNote. More than 3.2 million eNotes have been registered since it began. More than 300 originators participate. So do over 50 warehouse lenders and more than 40 correspondent investors.

The reason it belongs in a capital markets guide is transfer friction. If you want eNotes to move cleanly from warehouse to investor, every party in the chain needs registry participation. Ask your warehouse lender and your top buyers before you promise anyone a digital delivery date. The registry is fine. The gaps are always at the edges.

Servicing retention changes which system you need

Retain or release is a balance sheet decision, and it drives your software requirements more than most buyers expect. If you release, the all-in price is the only number that matters. Fannie's servicing-released execution shows all-in pricing and funding with your chosen servicing buyer. Freddie runs the same idea through Cash-Released XChange. It needs separate approval and about three business days to switch on. Freddie can withdraw the approval if you do not deliver within six months.

If you retain, you need loan-level MSR valuation feeding the same best execution comparison as the cash bid. Optimal Blue includes loan-level MSR valuation inside CompassEdge. MCT sells MSR valuation and analysis alongside its hedging service. A best execution engine that cannot value retained servicing is not comparing your real choices.

Non-agency and the newer venues

MAXEX solves a counterparty problem rather than a workflow one. It operates as a single clearinghouse counterparty between sellers and a pool of institutional buyers. Programs cover jumbo, conforming investor and second home, non-QM, and DSCR. In February 2026 it announced work with Tradeweb to widen institutional access to residential private credit. If your non-agency execution today means four separate approvals and four sets of reps, that is the pitch.

Figure Technology Solutions sits in this ranking because of Figure Connect, its loan marketplace. Its first quarter 2026 results, published 11 May 2026, put marketplace volume at $2.9 billion. That was up 113 percent year on year. Figure Connect accounted for $1.6 billion of that, up 237 percent, with 387 active partners at quarter end. Those are impressive growth numbers. They are also mostly home equity flow. Judge it as a distribution venue, not a replacement for your hedging stack.

What to test before you sign

Nobody in this category publishes a price list, so build your own comparison. Ask for the fee in basis points and in dollars per loan at your actual volume. Ask what happens to the fee in a year when volume halves. Then ask how the number changes if you add a channel or a second agency outlet.

Run the pilot on a real week. Feed the system a pipeline you already hedged and compare the position it produces to the position you ran. Give it a bid tape you already sold and check whether its best execution answer matches the outcome you got. Ask for a mid-month agency change to be shipped while you are watching. Optimal Blue cites a client that went live in under 60 days. MCT says an existing client can switch on its analytics product in a day. Hold both to those numbers in writing.

The last question is the one people skip. Ask who answers when the mark looks wrong at 8:45 in the morning. Get a name, a phone number, a response time, and an escalation path in the contract. Here the software is only as good as the person behind it on a bad day.