Mortgage Accounting Software

What decides this purchase

This purchase is decided by whether your general ledger understands a loan as a record. Everything else follows from that answer. A mortgage banker's books are not a services company's books with different account names. Loans sit on the balance sheet as inventory. They fund on a warehouse line. They sell at a gain or a loss. Some leave behind a servicing asset. A ledger that treats each funding as a bank transaction loses the thread on day one.

Production speed barely matters here. Accounting software does not close loans. Nobody in a branch waits on it. The weight sits on depth instead. What the package knows about mortgage banking decides whether your close takes four days or fourteen. Integration comes next, because the ledger is downstream of everything. Price and training matter least. A cheap ledger that produces a wrong branch P&L costs more than it saves.

The 90-day clock sets the requirement

Approved sellers and servicers work against one hard annual deadline. Fannie Mae's Selling Guide, at A4-1-02, requires audited financial statements within 90 days after fiscal year end. Ginnie Mae's MBS Guide sets the same 90-day window for issuers. HUD sets it too, at 24 CFR 202.5(g), for approved mortgagees. The date does not move because your close ran long.

Fannie Mae also names the pieces. The package needs a balance sheet, an income statement, a statement of retained earnings, and the related notes. It needs an independent public accountant's opinion. It needs GAAP presentation and prior-year comparatives. If the statements are consolidated with a parent, your entity has to be readable on its own. That last requirement quietly rules out a lot of chart-of-accounts designs.

Net worth and liquidity are computed from your ledger

The eligibility tests are formulas, and the inputs come out of your accounts. Fannie Mae requires adjusted net worth of at least $2.5 million. On top of that sits 25 basis points of Fannie and Freddie first lien servicing UPB. Ginnie Mae servicing carries 35 basis points. Non-depositories also hold an adjusted net worth to total assets ratio of 6 percent. Allowable liquidity runs off servicing UPB as well.

Ginnie Mae issuers face their own version. The MBS Guide sets single-family net worth at $2.5 million plus 35 basis points of effective outstanding obligations. Liquid assets run to the greater of $1 million or 10 basis points of outstanding single-family MBS. FHA approval under 24 CFR 202.5(n) starts at $1 million of net worth. It adds one percent of volume above $25 million, capped at $2.5 million.

None of that is difficult math. The work is producing the inputs on time and defending them later. A ledger that cannot split servicing UPB by agency turns every quarter into a manual exercise.

Loan-level detail is the whole argument

Ask a mortgage-specific vendor what they sell and the answer is the loan record. Loan Vision states it plainly on its platform pages. Every transaction ties to a specific loan from the moment it lands. Advantage Systems leads with the same idea for AMB, under the heading of loan level detail. The reason is cost to originate. Without loan-level costing you cannot say which product line earns money.

Gain on sale is the test case. A loan funds, sits in inventory, then sells to an investor. The gain is proceeds against carrying amount, adjusted for fees, premiums, discounts and any servicing retained. Spread that across a ledger with no loan dimension and the number becomes a journal entry someone typed. An auditor will ask how it was derived. Spreadsheet is a weak answer.

Warehouse lines are not bank accounts

This is where general packages break first. A warehouse facility is a revolving line with loans pledged against it. Each draw belongs to a loan. Each payoff belongs to that same loan, days or weeks later. Interest accrues per loan, not per line. Reconciling the facility means matching the warehouse lender's report to your loan detail.

Loan Vision describes warehouse lines as master records tied to the loans funded against them. That design is the point. If your ledger holds one balance per facility, someone rebuilds the detail in Excel every month. That reconciliation is also the one an auditor pulls first. Aged items sitting on it are how a clean opinion turns into a finding.

Branch P&L, and the reason it goes wrong

Most lenders allocate by branch and most do it badly. The failure is rarely the allocation logic. It is that revenue and cost arrive from different systems on different days. Commissions come from the LOS or a payroll file. Fee income comes off the closing disclosure. Rent and licensing sit somewhere else again. Corporate overhead comes from a table someone maintains by hand. Without a shared loan key, the branch P&L is an estimate.

Loan Vision publishes a customer account of exactly this fix. Mortgage 1 combined its branch accounts into one system. It could then run a P&L for any individual branch. That outcome is what you are buying. Not a report writer. The ability to hand a branch manager a number they cannot argue with.

MSRs and the election you already made

If you retain servicing, the ledger inherits an accounting policy choice. FASB Statement 156, now carried in ASC 860-50, requires servicing assets to be recognized at fair value on day one. After that you elect a subsequent method by class. The amortization method spreads the asset over servicing income and tests for impairment. The fair value method remeasures each period through earnings. Once fair value is elected for a class, that election cannot be reversed.

Neither ranked product markets itself as a valuation engine. Loan Vision's platform pages make no MSR claim at all. The valuation comes from a specialist model or an outside firm. What the ledger owes you is the booking and the roll-forward behind the disclosure. Ask how the monthly entry arrives and who owns the file it comes from.

TRID fees land in accounting, not compliance

Fee tolerance reads as a compliance rule and settles as an accounting one. Regulation Z at 1026.19(e)(3) sets zero tolerance on some charges and a 10 percent aggregate limit on others. When the limit is breached, 1026.19(f)(2)(v) requires the refund and a corrected disclosure within 60 days of consummation. Someone has to cut that check and book it against the loan.

That works only if closing disclosure fees import at loan level and reconcile to what was actually collected. Lenders tracking cures in a spreadsheet find them at quarter end. By then the 60 days are gone. It is a small feature with a large downside.

The quarterly filing nobody staffs for

State licensing adds a rhythm the annual audit does not. Licensed companies file a Mortgage Call Report through NMLS each quarter. It carries loan activity by state plus a company-level financial condition section. GSE-approved sellers and servicers file the Expanded form, as do Ginnie Mae issuers. Everyone else files Standard.

The financial condition section pulls straight from your books. If the ledger cannot produce it without rekeying, four filings a year become four fire drills. Ask each vendor to show you the report that feeds it.

Why only two products are ranked

This category is thin, and the reason is structural. Mortgage accounting serves a small buyer set, and much of that set stays on a general package. Accounting is a cost center, so vendors chase origination and servicing instead. Building a mortgage ledger also means building a ledger first. That is a heavy lift for a narrow market. So two names carry the category.

Loan Vision scores 4.8 here. Advantage Systems scores 3.9. Both do the mortgage-specific work that general packages skip. The gap between them is currency rather than concept. Advantage Systems' own FAQ page states its client count as of September 30, 2016. The same page gives its server requirement as 64-bit Windows Server 2012 R2 or newer. It describes loan integration as importing funding transactions from comma-delimited files.

That is a working design, and firms run on it today. It is not a modern integration story. Say your LOS exports clean files and your team knows AMB. Then the case for switching is weaker than the score gap looks.

Loan Vision and the Business Central question

Loan Vision runs on Microsoft Dynamics 365 Business Central and says so openly. It presents itself as a Microsoft Certified Partner for that platform. There is a budget consequence. Microsoft publishes Business Central list pricing at $80 per user per month for Essentials and $110 for Premium, paid yearly. Team Members seats are listed at $8. Those seats sit underneath the mortgage layer.

Ask what the platform buys you beyond the mortgage modules. Dimensions for branch reporting come from Business Central. So do the base ledger and the approval workflows. Loan Vision names Concur among its integrations and supports API and file-based imports from an LOS. It does not publish a list of named LOS partners. Make them name yours in writing before you sign.

When a general package is still the right answer

Plenty of lenders run QuickBooks, Sage Intacct, NetSuite or Microsoft Dynamics and pass their audits every year. The honest test is volume and structure, not vendor preference. A single-branch originator selling servicing released, funding on one facility, does not need a mortgage ledger. QuickBooks Online publishes list pricing, with its Advanced plan at $340 per month for 25 users.

Price is not the real comparison. The comparison is the cost of the people who patch the gap. Two analysts spending a week a month in Excel is a real line item. It never appears in the software budget.

The break point shows up in a few places. Branch count above a handful. Retained servicing of any size. More than one warehouse facility. Correspondent or broker volume running alongside retail. Hit two of those and the spreadsheet layer stops being temporary.

Where the money actually goes

Neither ranked vendor publishes a price. Advantage Systems states that cost depends on functionality and user count, with implementation billed hourly. Loan Vision publishes nothing on price. Expect four line items: platform seats, the mortgage layer, implementation, and a data conversion. Budget the conversion honestly. Loading historical loan-level detail into a new ledger is where these projects slip.

The return shows up in days rather than dollars. A close that drops from twelve days to six frees the same team for reporting nobody currently gets to. Loan Vision publishes customer accounts of that shape, including a 30 percent reduction in month-end close at Banksouth Mortgage. Treat those as vendor claims. Then ask for the reference call and check them.

What to test before you sign

Run the demo on your data, not theirs. Give the vendor one month of funded loans and one warehouse statement. Ask them to reconcile it in front of you. Then ask for the branch P&L that falls out of it. That is slower than a canned demo. It is also the only version that predicts what you will get.

Four questions separate these packages. How does a loan move from the LOS into the ledger, and what happens when a field changes after funding? How are commissions calculated, and how are they reversed when a loan dies? How do a warehouse draw and its payoff tie back to one loan? What does the auditor receive at year end, and in what format?

Get those four answered by a person who will still be on the account in year two. The demo is not where these deals go wrong. Implementation is.