The following is a guest post from Azhar Jaffri, co-founder and finance lead at Korven Consulting. Opinions are the author’s own.
In his Oct. 15 opinion piece, Duncan Young makes the case for updating cash forecasts every month rather than relying on an annual budget that quickly becomes outdated. I agree with the discipline, but there is a condition that deserves more attention: The reforecast is only as reliable as the close underneath it.
If the previous month still contains unreconciled accounts, missing accruals, late payables or cutoff errors, the forecast is not starting from a clean position. Updating the model more often does not solve that problem. You are updating the error.
What a messy close really looks like
A messy close does not always look dramatic. The P&L may look reasonable, the bank account may look healthy and revenue may be close to plan. The problems often sit underneath the headline numbers: An unreconciled account, a late supplier invoice, an old clearing balance, a transaction recorded in the wrong period or a one-time expense being treated as part of the normal run rate.
I saw this in one engagement where we identified a $64,880 supplemental property-tax charge. The expense was real, but it was not recurring. If management had simply rolled that month’s expense pattern into the forecast without understanding what created the charge, future cash needs would have been overstated.
That is why the close has to answer a basic question before a number becomes a forecast assumption: What created this balance, and can we support it?
Three habits that make a forecast trustworthy
1. Close by business day 10, with reconciliations complete
For many companies, business day 10 is a reasonable target. The exact deadline can vary, but “closed” should mean more than the accounting period has been locked.
Bank and card accounts should be reconciled, material balance-sheet accounts should tie to supporting schedules, receivables and payables should be reviewed and significant payroll, tax, debt and other liabilities should be understood. Any material reconciling item that remains open should be documented, assigned and communicated to the people relying on the forecast.
A fast close is useful. A fast close that needs major corrections two weeks later is not.
2. Use an accrual and cutoff checklist
Not every cost arrives neatly before month-end. Freight, professional services, utilities, bonuses, contractor costs and other obligations may belong to the period even when the invoice arrives later.
A simple checklist forces finance to ask what has been incurred but not recorded, what belongs to the next period and what liability exists today even though the cash will leave later. That distinction matters because accounting recognition and cash timing are not the same thing.
The close should establish the obligation. The 13-week forecast should establish when the cash is expected to move. Mixing the two creates bad accounting and bad forecasting at the same time.
3. Review forecast versus actual every month
A forecast should not just be refreshed. It should be challenged.
Compare expected cash receipts and payments with what actually happened, then explain the material differences. Were collections late? Did inventory purchases run higher than expected? Did payroll or taxes hit in a different week? Was the variance caused by timing, volume, pricing, hiring or simply a bad assumption?
The goal is not to eliminate every variance. The goal is to learn from it. If customers consistently pay seven days later than the forecast assumes, update the collection assumption. If a category repeatedly misses because the business model changed, stop treating the miss as an exception.
The close should feed the 13-week cash forecast
A 13-week forecast should begin with reconciled cash, not simply the balance someone sees when logging into the bank. From there, finance should layer in expected collections, open payables, payroll, taxes, debt service, recurring operating costs, capital expenditures and known one-time payments.
The bank balance alone does not tell management how much cash is truly available. A company can have $500,000 in the bank and still face a liquidity problem if $620,000 of obligations are due over the next six weeks and expected collections are not enough to cover the gap.
Each forecast week should show opening cash, expected inflows, expected outflows and closing cash. The closing balance rolls into the next week. As actual results replace forecast weeks, the model should extend forward so management always has a rolling 13-week view.
This is where the close earns its value. It gives the forecast a starting point management can defend.
What CFOs should measure
Three measures tell you a great deal about whether the process is improving.
1. Days to close: How long after month-end does finance produce numbers management can actually use? If the close routinely takes three weeks, the company is spending much of the next month looking backward.
2. Post-close adjustments: How many material corrections are made after the financials are considered complete? A high number usually points to weaknesses in reconciliations, accruals or review.
3. Forecast accuracy: Compare forecast to actual for ending cash, collections, payroll, major operating expenses and large cash commitments. Do not only ask whether the forecast missed. Ask why it missed and whether the assumptions improve as a result.
A faster close is not automatically a better close, and forecast accuracy by coincidence is not a strong forecasting process. The objective is timely information that management can explain and use.
If you fix only one thing this quarter, fix the close before making the forecasting model more sophisticated.
Before buying another forecasting tool or adding more scenarios, make sure finance can explain the opening balances, reconcile material accounts, record obligations in the correct period and review forecast-versus-actual differences consistently.
A simple 13-week forecast built on reconciled numbers is more useful than a sophisticated model built on questionable actuals. The forecast is a promise built on the books. Fix the books first.
