Skip to content
All free toolsMenu

Type a task or a name, for example “remove background”, “PDF” or “P0420”.

Money & home·7 min read

How to compare two mortgage models without mixing assumptions

Put the offers on the same foundation before comparing their model results. A difference caused by an omitted cost is not evidence that one lender charges less for the same loan.

The short answer

Gather both offers, identify the loan amount and repayment method, then open “Compare offers” in Mortgage Studio. Keep the original documents beside the model and record every omission. This guide explains a comparison workflow, not a recommendation to borrow.

Open the free tool

Prepare one comparison sheet before entering offers

Write down what both documents describe: the proposed loan, currency, repayment approach and dates. Give each document a clear name such as “Offer A, received Monday” rather than relying on a lender logo alone. If a revised document arrives, preserve the earlier version with its date. Otherwise a changed fee can be mistaken for an error in your calculation.

Choose one question for the model. You might want to compare the same entered amount over the same term, or inspect balances at the end of the fixed-rate period. Those are different comparisons. Record the question at the top of the sheet so a monthly-payment difference is not later presented as a complete judgement about the offers.

Separate common inputs from offer-specific inputs

The property price and your intended contribution may stay the same while the loan terms differ. Identify those shared assumptions first. Keep buying costs separate from money available to reduce the price. If the savings input includes both, record that fact rather than comparing it with another version that includes only the down payment.

Do not force both documents to show the same loan by silently removing an expense from one. If you deliberately create an equal-amount model, label it as your comparison scenario. The document supplied by the lender remains the source for its own stated amount. A changed assumption can be useful, provided it is visible and not passed off as the original offer.

Enter the rate that the model requests

Mortgage Studio's plan asks for a nominal annual rate. Copy the corresponding figure from the document, not another percentage that happens to appear nearby. Keep a fee-inclusive rate in your comparison notes with its original label. If the document's terminology is unclear, ask the issuer to identify the appropriate figure before creating a model that appears precise.

For the US documents discussed by the CFPB, the official comparison guidance covers more than the headline rate. That source was checked on 22 October 2026. Its form names should not be presented as a universal requirement in another country. The practical task here is to preserve what each document actually includes and to ask about unclear differences.

Build the first plan with traceable assumptions

Start in “Payment & plan”, select the applicable country and enter the figures that belong to the chosen scenario. Record any starting value you replace. A country preset is an aid to assembling a model; it does not confirm the charges for a particular property or transaction. Keep unconfirmed inputs explicitly marked in your comparison sheet.

Read the resulting payment and schedule with those limits in view. Then use “Add to comparison” to carry that scenario into the comparison workspace. Give it the same name as your sheet. If you change the source document later, update or replace the related scenario rather than leaving two identically named cards that represent different assumptions.

Add the other offer without changing the question

Open “Compare offers” and use “Add offer” where needed. Check the common comparison amount and duration before filling the offer-specific entries. Keep the names and cost meanings consistent across cards. An omitted charge should be described as omitted or unresolved, not entered as zero merely because the page does not have the information.

The model can compare only the information its fields represent. Keep contractual details that have no matching field in a separate note. For example, a document may describe a condition whose effect cannot be reduced to one input. Do not invent a numeric equivalent simply to remove the note from the comparison. The unresolved condition belongs in the conversation with the issuer.

Read the comparison at the horizon you chose

Inspect the duration used for a displayed cost comparison and distinguish it from the full loan duration. A result over a selected period does not describe every later payment. If a rate is fixed only for part of the model, a curve beyond that boundary rests on assumptions. It does not establish the rate available when that period ends.

Use the remaining balance and the schedule to understand what the model leaves outstanding at that point. Two scenarios with similar early payments can leave different balances under their entered terms. Describe those differences with the input assumptions attached. A label such as “cheapest” on a calculated table should not be repeated as a conclusion about every contractual or personal consideration.

Resolve mismatches before trusting a tidy table

If a lender payment and the model differ, list possible inclusions first: the rate definition, timing, insurance or another item the document counts. Check the units and whether a field expects an annual or monthly amount. Do not alter a rate until the outputs happen to match while leaving the true source of the difference unexplained.

Keep a question for each unresolved mismatch and ask for an explanation tied to the document. A useful note states which line and which assumption disagree. The aim is a reproducible comparison, not a screenshot that suppresses every uncertainty. If a charge cannot yet be entered reliably, preserve the model as provisional rather than presenting its totals as complete.

Preserve a record another person can follow

Download the schedule as CSV or use “Print / PDF” for a readable copy. Keep the offer documents with your comparison notes, including the chosen horizon and excluded conditions. If you share a scenario link, review the figures it conveys and choose its recipient deliberately. Local browser saving does not control copies that you have already forwarded.

Before a later update, duplicate your notes with a date and identify what changed. Replace one assumption at a time when investigating an unexplained difference. This leaves a clear route from document to model to question. The final decision still requires understanding the actual agreement; the calculator cannot read its full terms or assess whether the commitment suits you.

Questions you might still have

Should I enter the APR as the nominal rate?

Use the rate definition requested by the field. Keep APR or another fee-inclusive measure under its own label in your notes, and clarify ambiguous document wording before calculating.

What if the two offers use different loan amounts?

Either preserve that difference explicitly or create a separate equal-amount comparison model. Do not describe changed figures as though they were the original documents.

Can an unknown fee be set to zero?

Only if zero is actually established. Otherwise keep it unresolved and explain that the model excludes it, rather than turning missing information into a favourable amount.

Does a lower model payment settle the choice?

No. The comparison describes entered assumptions and included fields. Duration, remaining balance and conditions outside the model still need to be understood.

Why keep older offer versions?

A date and version let you explain which figures produced the saved model. A later revision can otherwise make an accurate older result look like an unexplained mistake.

What should I send when asking about a mismatch?

Identify the document line, the input used and the model output that differs. Include the inclusion or timing question instead of changing unrelated numbers until they match.

Sources & useful details

These guides are written from the current LoschaZ tools and the sources listed above. Examples are illustrations, not tests on every device or promises of an identical result.