Skip to content

New: How We Design Technology - our seven-layer architecture model, explained in full. See the architecture

general

FFIEC Business Continuity: Proving Restores to Examiners

11/08/2026

The request list for next month's exam arrives, and one line stops you: evidence of backup and recovery testing performed since the last examination. You have a year of green backup job reports. You also know the last full server restore happened during a hardware failure eighteen months ago, and nobody wrote down how long it took.

Examiners applying FFIEC business continuity expectations want proof that your institution can recover, and a successful backup job only proves that data was copied. The good news is that the proof is within reach. A small set of well-chosen restore tests, measured carefully and recorded on one page each, answers the question examiners are asking and tells your board something useful at the same time.

At a glance

  • The FFIEC Business Continuity Management booklet treats tests as a way to verify the performance and reliability of your systems, and it expects results and identified issues to reach the board.
  • Set recovery objectives with the business first, then design restore tests that reflect the disruptions you're most likely to face.
  • Measure real restore times, record each test on one page, and track every gap until a retest shows it's closed.

What the FFIEC business continuity booklet expects

The FFIEC revised the Business Continuity Management booklet of its IT Examination Handbook in November 2019, replacing the 2015 Business Continuity Planning booklet. The OCC, FDIC, Federal Reserve, and NCUA each announced the revision. The booklet helps examiners judge whether management adequately addresses risks to the availability of critical financial products and services. The FDIC's announcement noted that the changes don't impose new requirements, so the shift is mostly in emphasis.

That emphasis is on resilience, with business continuity treated as ongoing management work. The booklet's components include strategies and plan development, training, exercises and tests, maintenance and improvement, and reporting to the board.

The booklet draws a useful line between the two kinds of validation. An exercise involves people and processes and checks one or more parts of the plan, such as a tabletop discussion of a ransomware scenario. A test is a type of exercise meant to verify the quality, performance, or reliability of system resilience, including whether a system's recovery time objective is acceptable. Restore testing falls squarely in that second category.

Agree on recovery objectives before you test

A test needs a target. Two terms do most of the work. The recovery time objective (RTO) is how long a system can be down before the disruption causes real harm to members or operations. The recovery point objective (RPO) is how much recent data you can afford to lose and re-create, measured in time. The booklet places both inside the business impact analysis, alongside maximum tolerable downtime.

Set these with the people who run the business. Ask the retail operations lead how long branches can work without the systems that support teller activity. Ask lending how much of a day's loan file activity they could rebuild from paper. The answers will differ by system, and that's the point. Your online banking connections, loan origination platform, email, and departmental file shares each deserve their own targets, ranked by how quickly their absence affects customers.

Design restore tests around real disruptions

A single annual test of one file server leaves most of your recovery path unproven. Choose a handful of scenarios that match the disruptions you're most likely to face, and rotate through them during the year.

A single file or folder. The most common request. It proves retention settings and the basic restore process.

A user mailbox or OneDrive. Shows whether your Microsoft 365 backup can bring back a departed employee's data or a mailbox someone damaged.

A full server. Measures the hardest piece: rebuilding a whole system, including the operating system, applications, and configuration.

Systems that connect to third parties. Your core processor and other providers run their own recovery. Coordinate with them so you know which interfaces, file transfers, and credentials have to come back on your side, and ask for their recent test results.

A ransomware-style recovery. Restore into an isolated environment from copies an attacker couldn't have changed, with your directory and privileged accounts rebuilt first. This shows whether you could recover if your production network couldn't be trusted.

Run the scenarios that carry the most risk more than once a year, and test again after major changes such as a core conversion, a new backup product, or a move to Microsoft 365.

Measure what actually happened

Start the clock when someone decides a restore is needed, and stop it when a business user confirms the system works. That window includes the steps that usually take longest: finding the right restore point, getting approvals, locating credentials, moving data across the network, and checking the result. A backup tool's job timer measures only the middle of that process. Note who had to be reached during the test and how long each wait lasted.

Record the age of the restore point too. If the newest clean copy was taken the previous night, your actual recovery point for that scenario was about a day, whatever the policy says. Compare both numbers with the objectives the business set. A test that misses its target still has value, because it tells you where to invest before a real outage does.

A one-page record examiners can follow

Each test deserves a short record that someone outside IT can read. Keep it to one page and keep the format consistent so results can be compared over time.

  • Scope and scenario
    • What to record: The system, the data, and the disruption being simulated
  • Objectives
    • What to record: The RTO and RPO the business set for this system
  • Date and participants
    • What to record: Who ran the test and which business user validated the result
  • Restore point used
    • What to record: Where the copy came from and how old it was
  • Measured results
    • What to record: Actual time to recover and actual data loss, compared with the objectives
  • Validation
    • What to record: How the business user confirmed the system and data were usable
  • Issues found
    • What to record: Anything that slowed or blocked recovery
  • Follow-up
    • What to record: An owner and due date for each issue, plus the planned retest

Keep the records with your business continuity documentation, and file any supporting evidence, such as screenshots or logs, alongside them.

Report results to the board

The booklet expects the board to receive a written presentation covering the business impact analysis, risk assessment, continuity plan, exercise and test results, and identified issues. It also expects board minutes to show real discussion and approvals, including credible challenge from directors.

Write that report for directors who don't run backups. Show which critical systems were tested, whether each met its objectives, what gaps were found, and who is fixing them by when. One honest missed target with a credible remediation plan builds more confidence than a page of uniform success rates.

Follow up until the gap is closed

The booklet also addresses actions after exercises and tests, and examiners will look for evidence that problems found in testing were fixed. Track each issue with an owner and a date. Update the continuity plan when a test shows a procedure is wrong. Then retest the same scenario, because the retest is what proves the fix worked.

Over a year or two, this rhythm gives you a clear record: objectives agreed with the business, tests that reflect real risk, measured results, and gaps that closed. That's the material an examiner can follow from the first request to the final page.

At CTP, we help banks and savings institutions build recovery paths around agreed objectives and document the testing that proves them. Learn more about managed IT for financial institutions, or schedule a conversation before your next exam cycle.

Sources and further reading

Ask about something important for you.

Curious about our take on an industry or technology topic? Let us know.

Contact Us