The UAE Central Bank Raised Its Base Rate to 3.90% on 17 September — the 5 Hiring Decisions I Regretted the Last Time Money Got Expensive in Dubai, and What I Am Doing Differently This Week

At 22:16 on Wednesday 16 September, Gulf News ran the line every Dubai finance director had been waiting for since the Fed statement dropped: “UAE Central Bank raises base rate to 3.9% after US Fed rate hike”. The Central Bank of the UAE lifted the Base Rate on its Overnight Deposit Facility from 3.65% to 3.90%, effective Thursday 17 September, and kept its standing credit facilities at 50 basis points above that. It was the mechanical consequence of the dirham peg: a few hours earlier the Federal Open Market Committee had raised the federal funds target range by 25 basis points to 3.75–4.00%, on a 12–0 vote, the first increase since 2023. I ran delivery for two Dubai software companies through the last tightening cycle, when the CBUAE base rate went from 0.15% in March 2022 to 5.40% by July 2023. I got five hiring decisions wrong in that period. This is what they were, and what I am doing differently this week.

What Happened on 16 and 17 September, in the Fed’s Own Words

The FOMC statement is short and unusually blunt. “Inflation remains elevated,” it says. “Today’s policy action will support a timelier return to the Committee’s 2 percent goal.” On the labour market: “Job gains have kept pace with the workforce, and the unemployment rate has changed little.” On growth: economic activity is “expanding at a solid pace”, with domestic spending described as resilient and productivity growth as strong. That combination, solid growth plus sticky inflation, is the recipe for more than one hike. CNBC’s read-out of the projections puts officials’ year-end rate between 4.1% and 4.4%, and markets are pricing at least one further 25-basis-point move before the end of 2026.

For the UAE the transmission is immediate. The National’s explainer of 16 September walks through the consumer side: variable-rate mortgages reprice, personal loans and card rates rise, savers finally get something for their deposits, and EIBOR, the benchmark that most business facilities in the country are priced off, is expected to move up. Nobody in that coverage wrote about engineering headcount. But headcount is where a rate move shows up first inside a software company, because payroll is the largest line, and the hiring plan is the only large line a founder can change in the same quarter.

Why a Quarter Point Hurts a Startup More Than It Hurts a Bank

A bank earns more on a rate rise. A software company with a revolving facility priced off EIBOR, or venture debt with a floating coupon, or simply an investor who models the value of future revenue off the risk-free rate, pays more for every month of runway from the day the rate changes. The arithmetic is unforgiving: if you have 18 months of cash and your facility reprices by 25 basis points, the direct cost is small; the indirect cost, that your next round is now priced against a higher hurdle and your investors want the burn lower before they lead it, is the number that reaches the hiring plan.

In 2022 I did not understand that sequencing. I treated the first 25-basis-point move in March as a footnote. By June the board wanted burn down by a third, and the only line I could move fast enough was headcount. What follows is the honest version of what we did next, and why most of it was wrong.

The 5 Hiring Decisions I Regretted in 2022–23

1. A blanket freeze instead of a re-tiering

We froze every open requisition in one email, including two backend roles on the only product line with signed enterprise pipeline. The product shipped nine months late. The senior engineers we kept spent that time on maintenance they were over-qualified for. A freeze feels decisive and is easy to explain to a board; it is also the hiring equivalent of turning off the lights in every room because one bill went up. What we should have done was rank every open role by how directly it shortened time to revenue, keep the top two or three, and pause the rest with an explicit review date.

2. Cutting the offshore pod first because it was the easiest line to cut

Our dedicated team in Karachi was on a 30-day notice contract, which made it the cheapest thing to stop, so we stopped it. That was backwards. The pod was our lowest cost per engineer and our most elastic capacity; the expensive, inelastic cost was the Dubai-employed team with visas, insurance and end-of-service accruing whether or not they had work. Cutting the elastic line to protect the inelastic one is the classic mistake of a rate cycle, and I have watched three other Dubai companies make it since. The right order is the opposite: hold the variable-cost capacity, and use the notice-period flexibility as the buffer it was designed to be.

3. Twelve-month fixed-fee vendor contracts with no rate clause

In Q1 2022 we signed two outsourcing agreements at fixed monthly fees for twelve months, with no review clause. By Q4 the vendors’ own costs had moved and so had ours, but the contract could not. One vendor quietly downgraded the seniority of the people on the account to protect its margin; we did not notice for four months. The fix is not shorter contracts, which cost you continuity; it is a 90-day review clause tied to an objective trigger, a named senior on the account, and an SLA that specifies who is doing the work, not just what gets delivered. Our piece on SLA clauses for outsourced Node.js work in the UAE has the language we now use.

4. Leaving contractors on a 30% premium for 14 months

Three of our best engineers were UAE-based contractors on freelance permits, billing day rates that worked out at roughly 30% above what the same people would have cost as employees. We had planned to convert them “once things settled”. Things did not settle for 14 months. The premium was justified for the flexibility in month one; by month six we had no intention of letting them go, so we were paying for an option we would never exercise. The conversion conversation is uncomfortable and takes a week; the premium is invisible and takes a year. The freelance permit route is excellent for genuinely temporary capacity and expensive for anything else.

5. Letting counter-offers set the salary bands

Through 2022 we matched three counter-offers to keep people, each time about AED 8,000 a month above band. By mid-2023 the market had softened, hiring had slowed across Dubai, and we had three people paid for a market that no longer existed, sitting next to peers paid at band. Two of the three left anyway within a year. A rate cycle is exactly when you should hold the band and lose the occasional person, because the replacement market gets cheaper as the cycle runs. We wrote up the discipline we now use in negotiating developer salaries and closing competitive offers in Dubai.

What I Am Doing Differently This Week

None of this requires panic. It requires doing in the first week what we did in the ninth month last time.

• Re-run the runway model at 4.25%, not 3.90%. The Fed has told you it is not finished. If your facility is EIBOR-linked, ask your bank for the reset date and put it in the hiring calendar.
• Re-tier every open role by revenue attribution. Three columns: shortens time to revenue, protects existing revenue, everything else. Fill the first column now, at band. Pause the third with a written review date in December.
• Move non-customer-facing capacity to variable cost. Internal tooling, data pipelines, QA automation and the second mobile platform are the work that belongs in a dedicated pod with a transparent rate card and 30-day notice, not on Dubai payroll with a two-year visa.
• Put a 90-day review clause in every new engagement, with a named senior on the account and the right to audit who is actually doing the work.
• Convert the contractors you already know you are keeping. Do it this month. The premium you are paying is the cheapest saving in the plan.
• Hold the salary band. Say so in writing to hiring managers before the first counter-offer arrives, because it will arrive in the week someone reads about the hike.

One Structural Note for Free-Zone and Mainland Companies

How much of this you can act on depends on your licence. A free-zone entity can generally engage contractors on freelance permits and add or remove pod capacity without touching its own visa quota; a mainland company with Emiratisation obligations has less room to reduce skilled headcount without moving away from its targets. If you are deciding where to incorporate, or whether to move, the mainland versus free zone guide and the DIFC fintech licence guide set out what each structure lets you flex, and the staff augmentation guide explains how the pod model sits alongside your own payroll.

If You Also Run a Team in Singapore

Singapore is not pegged to the dollar, and the Monetary Authority of Singapore steers the exchange rate rather than a policy rate, so the transmission there is slower and runs through SORA rather than a same-day base-rate move. The pressure on hiring plans is real all the same, and the decisions above translate almost line for line. Our Singapore colleagues wrote up how to hire during a tech downturn in Singapore and how to structure developer compensation packages there; both are useful reading for anyone running a two-city budget this quarter.

FAQ — The UAE Base Rate Rise and Dubai Developer Hiring