Let me concede the easy point. ING's call that the Reserve Bank of New Zealand will hike with CPI printing close to forecast is directionally plausible. Central banks hike when inflation holds near projection and growth prints firm. That is textbook. The pattern we keep seeing — across broker dashboards offering leverage from 1:400 at AvaTrade up through 1:2000 at Exness and FXTM, up to 1:3000 at FBS — is that beginners read "hike seen with CPI near forecast" as a trade entry. It is not one. It is a sentence about probability conditions, not about price. The math between the two is where retail accounts die.
The "Near Forecast" Fallacy
There is a pattern we keep encountering in reader inboxes. An analyst note crosses the wire. The reader latches onto the phrase "near forecast" and reads it as a green light. The grammar is doing more work than the economics.
Think about what "near forecast" actually means in probability terms. A central bank publishes a numerical projection. Say the projection is 2.9 percent year-over-year. "Near forecast" is a soft band — call it plus or minus 20 basis points around the number. That band contains three categorically different outcomes. A print that comes in at 2.7 percent is a modest downside miss that may delay the hike. A print at 2.9 percent is on the pin and raises hike probability. A print at 3.1 percent is a modest upside miss that could pull the hike forward or lift its magnitude. All three outcomes qualify as "near forecast" in casual English. All three price differently in NZD crosses.
The math is simple and brutal. If "near" means a 40-basis-point band, and the three outcome zones within that band have different implied forward paths, then the information content of the headline is roughly one-third of what the reader thinks they received. The analyst's sentence collapses a distribution into a point estimate. The reader then collapses the point estimate into a trade direction. Two layers of information loss, and the reader has not placed a single order yet.
What professional desks do with the same note is read it as a conditional. The hike probability becomes a function of where in the band the actual print lands, with path dependency on what the RBNZ said in its last Monetary Policy Statement about the symmetry of its reaction function. Retail does not do this. Retail reads the headline, opens a chart, and clicks buy on NZD/USD.
The Analyst Byline Shortcut
The second pattern is older than the forex retail industry itself. Readers treat the institutional byline as a substitute for the arithmetic.
ING is a credible shop. Its FX strategy team publishes often, is quoted on Bloomberg and Reuters, and sometimes gets rate calls right before the consensus moves. None of this means that an individual ING note, read in isolation, is a tradable signal. The pattern we see in reader behaviour is the compression of "ING argued, based on their forward curve model and their read of the latest MPS, that a hike becomes more probable if CPI prints near forecast" down to "ING says hike, buy NZD." That compression is where the error lives.
A proper read of any sell-side forecast involves three steps the retail reader usually skips. First, what does the house's forward curve already price? If OIS already has 15 basis points of hike priced into the next meeting, then ING saying "hike expected" is a confirmation of the base case, not a new signal. Second, what is the delta versus consensus? A Reuters poll of economists is public. If 70 percent of polled economists see a hike, ING is in the crowd. If only 30 percent do, ING is contrarian — and contrarian calls that print right move the price more than consensus calls that print right. Third, what is the house's own track record on this specific central bank? Not every FX desk is equally good on every currency. Specialist knowledge of RBNZ communication rhetoric is a distinct skill from eurozone rate-call accuracy.
The retail shortcut is to skip all three steps and treat "ING said" as the trade thesis. The desk's name is doing emotional work that the arithmetic refuses to do. We have seen the same shortcut applied to every tier-one analyst byline — Goldman, JPMorgan, HSBC, Deutsche, Barclays. The institution is credible. The reader's compression is not.
A central bank hike that was telegraphed six weeks ago is not a trade; it is a settled debt the market has already paid for.
The Already-In-The-Price Problem
Here is where the math starts to bite harder. Assume ING is right. Assume CPI prints on the pin. Assume the RBNZ hikes by 25 basis points on the next meeting date. The retail reader who went long NZD/USD on the back of the ING note then opens their broker platform expecting a reward. What they find is often a shrug.
The reason is not that forex markets are broken. The reason is that a widely-expected rate decision contains very little new information at the moment of announcement. The hike was priced into the OIS curve weeks earlier. The hike was discussed in the previous MPS. The hike was in the Reuters consensus. Every incremental hour between the forecast call and the decision date is an hour in which the market absorbs the probability and discounts the forward rate accordingly. By the time the Governor steps to the lectern, the directional move has largely already occurred — in slow, incremental repricing that the reader was not watching because they were waiting for the "event."
The academic literature calls this the expectations channel of monetary policy transmission. The practitioner version is simpler. Price moves on surprises. If the market expected a hike, the hike itself is not a surprise. The surprise would be a hold, or a 50-basis-point move, or hawkish forward guidance that was not previously signalled, or the opposite — a hike paired with dovish rhetoric that cuts the implied forward path.
For the retail reader long NZD/USD at 0.60-ish levels after reading the ING note, the asymmetry is uncomfortable. If the RBNZ delivers exactly what was expected, the pair might move 20 pips either way on noise. If the RBNZ surprises dovishly, the pair can drop 80 to 150 pips in minutes. The reader's trade has negative convexity against the thing they were betting on. Being right about the hike is worth little; being wrong about the surprise component is worth a lot. This is not a feature of the NZD specifically. It is a feature of every well-telegraphed central bank decision in every G10 currency since the Federal Reserve started publishing dot plots.
The Leverage-Compounded Reading Error
Here is where the broker data in front of us becomes relevant. Up to this point we have been describing a probability-reading error. The error is survivable at low leverage. It is not survivable at the leverage ratios the modern retail industry offers.
Run the arithmetic. A beginner opens a FBS account with the minimum one-dollar deposit, funded up to say two hundred dollars. The platform offers leverage up to 1:3000. On an NZD/USD trade, each micro-lot (1,000 units) at 1:3000 effective leverage represents an exposure of roughly 600 US dollars against account equity of 200. That is a notional-to-equity ratio of 3-to-1 before stop-loss sizing. A single standard lot — 100,000 units — at that same leverage would be 60,000 US dollars of notional against 200 of equity. The account is three hundred times over its own cushion. A 30-pip adverse move erases the entire deposit. The margin call fires before the trader has time to re-read the ING note.
The spread between brokers tells the same story at different magnitudes. Exness offers up to 1:2000 and a one-dollar minimum deposit. FXTM runs up to 1:2000 with a ten-dollar minimum. HF Markets caps at 1:1000. AvaTrade, the most conservative of the five, caps at 1:400 with a hundred-dollar minimum deposit. The common thread is that all five brokers allow a beginner, with less capital than a mid-range dinner costs, to open a position whose adverse move of less than one percent eliminates the account. The ING note never contemplated that reader. The ING note was written for institutional subscribers managing books where leverage is implicit in balance-sheet structure, not quoted at 1:3000 on a mobile screen.
The teardown of the broker landscape here matters because the two primary documents point in opposite directions and the retail reader does not know they are reading both at once. The analyst note implies a slow, probability-weighted expression of a view — the kind of position a hedged book might carry for weeks. The broker interface implies a fast, binary-outcome expression — the kind of position that resolves in hours or minutes under maximum leverage. The reader reconciles the contradiction by hitting the leverage slider at the top of the broker range and clicking buy. Both documents are "operative" in the sense that both exist on the trader's desk. Neither was designed for the use the trader is making of them.
There is a secondary math layer that deserves one line. The spread differences — 0.9 pip average on AvaTrade, 1.0 at Exness standard versus 0.1 on Exness Pro, 0.7 at FBS, 1.5 at FXTM, 1.2 at HF Markets — are second-order compared to the leverage math. A beginner lured by a 0.1-pip spread on Exness Pro who then uses the full 1:2000 leverage has saved perhaps a dollar on execution cost while taking a thousand-dollar risk on sizing. The marketing emphasis on tight spreads is a diversion from the arithmetic that actually ends the account.
So What Do You Actually Do
Read the analyst note as a conditional probability update, not as a trade entry. The sentence "RBNZ hike seen with CPI near forecast" is a sell-side house raising its subjective probability of a specific outcome. The appropriate response is to adjust your own prior accordingly if you have one, or — if you do not have an independent view on New Zealand rates and never did — to notice that you have no edge in the trade the analyst is implicitly describing. The absence of edge is a position. Flat is a position.
If you do have a prior, test it against the two cross-reference questions. Is this already in the OIS curve? Is ING in the consensus or against it? Both answers are findable in public data — the New Zealand short-end swap curve is quoted on every data terminal, the Reuters consensus poll is routinely summarised in free press coverage. If the hike is priced and ING is in the crowd, there is no trade in the headline. There may be a trade in the surprise component on decision day — which is a different trade, requiring different sizing and a different holding period, usually measured in minutes around the release.
If you are reading this on a broker platform that offers you 1:1000 or 1:2000 or 1:3000, treat the leverage slider as the single most dangerous control on the screen. Position sizing done against the maximum leverage the broker allows is not aggressive trading; it is a configuration error. The brokers in the grounding — AvaTrade at 1:400, HF Markets at 1:1000, Exness and FXTM at 1:2000, FBS at 1:3000 — are offering tools calibrated for a professional range of use. The retail misuse of those tools, pairing them with headline-driven directional bets on central bank meetings, is the industry's quietly observed failure mode. The analyst note did not cause it. The leverage slider did not cause it. The combination, in the hands of a reader who skipped the arithmetic, does.
Fieldnotes. The ING note that triggers reader questions is almost never the longest or most technical one the desk has published that month; it is the one with the shortest, most quotable headline. The Reuters consensus poll on RBNZ decisions is public but rarely cited in the broker platforms' own research tabs. Four of the five brokers in our grounding allow leverage settings high enough to vaporise a two-hundred-dollar deposit on a sub-one-percent adverse move; the fifth, AvaTrade at 1:400, is the outlier in the sample, not the rule. Every retail trader we have ever seen trade an RBNZ decision at maximum leverage has described the ING or Goldman or JPMorgan note that convinced them, by name. Not one has described the OIS curve as of the morning of the trade.