Michael Hartnett's note landed with the clarity his desk tends to prefer: risk-off stays bid until the dollar prints its cycle high, and not a session earlier. The question we keep getting is not whether the thesis is correct — it is what a retail trader is supposed to *do* with it, given that "wait for DXY to peak" is a macro instruction, not a position. So it depends. It depends on how you are funded, how you clear, and which platform your broker actually quotes on. We will walk through three composite scenarios — hypothetical readers only, no fabricated interviews — and show the math on each, drawing on the broker data our desk has verified: AvaTrade's 400:1 cap, Exness's 2000:1 leverage with 0.1-pip Pro spreads, FBS's 3000:1 and 1 USD minimum, FXTM's rupee accounts, and HF Markets' DFSA-regulated 1000:1 book.
Before we go further, the concession: Hartnett is right that regimes do not reverse until the funding currency tops. That part of the argument is clean. What we want to dismantle is the idea that the thesis translates into "stay in cash and wait." It does not. It translates into a very specific set of position-management decisions, and those decisions look different depending on who is reading the note. That is the whole point of walking through the three scenarios.
Scenario 1: The Micro-Lot Carry Grinder Holding USD Longs
Picture a trader who has been running a boring long-USD book for the better part of a year — small size, held through swap, no fireworks. Let us say the account sits at HF Markets, chosen because the DFSA tier-1 regulation reassures them and the 1000:1 leverage cap is enough headroom to run a 20:1 effective exposure without feeling cramped. The 1.2-pip average spread on EUR/USD is wider than they would like, but the HFM App makes position monitoring easy and withdrawals clear in a day. This is not a scalper. This is somebody who sized a position around a carry thesis and now has to decide whether Hartnett's call changes anything.
Here is the math, and it is the whole point of the scenario. Let us say the account equity is 10,000 USD. The trader is short EUR/USD at 0.25 lots — that is 25,000 EUR of notional, roughly 27,250 USD at a 1.09 reference rate. Required margin at 1000:1 is 27.25 USD. Free margin: 9,972.75 USD. Daily swap credit on short EUR/USD is positive — call it 2.80 USD per night on 0.25 lots in a typical rate differential environment. Over 30 nights, that is 84 USD, or 0.84% of equity per month, pure carry. Compound that to the back half of 2026 and it is a 4.2% six-month yield before any price move.
Now layer the thesis. If Hartnett is right and DXY grinds higher before peaking, the carry trade keeps paying and the mark-to-market gets better. If he is right that it peaks and reverses sharply, the carry still pays until the day of the turn, but the mark-to-market gives back fast. The question is not whether to exit. The question is where to put the stop.
Here is where platform matters. On MT5 — which HFM supports — this trader can run a trailing stop that steps 15 pips at a time, server-side, and does not need the terminal running. On the HFM App alone, server-side trailing is available but the step granularity is coarser. On a cTrader-based broker they would get smoother execution, but HFM does not quote cTrader. So the decision is: use MT5, set the trailing 60 pips below the current market, let the server manage it, go to sleep. The thesis says "stay in until DXY tops." The execution says "let a trailing stop decide when top is in."
For this trader, Hartnett's note changes nothing about the position and everything about the exit discipline. The platform decision — MT5 for server-side trailing — is the operational translation of the macro call.
Scenario 2: The Weekend Scalper Running MT4 on a Promo-Funded Account
Now imagine a different reader. Picture a trader who opened an account with 100 USD of their own money plus one of the historical no-deposit promotions — the kind that structured the retail landscape before CySEC's 2018 marketing restrictions cleaned up the EU side and ASIC followed in 2020. Pre-restriction, 30-100 USD welcome credits were common; post-restriction, the model survives mainly through non-EU licenses, and the wagering-requirement math is why most never convert.
Let us say the trader is at FBS — the 1 USD minimum deposit and 3000:1 leverage are what pulled them in — and they trade on MT4 because that is where their indicator stack lives. Average EUR/USD spread of 0.7 pips on the standard account, 0.0 pips on Pro but with commission. They scalp Sunday-evening gap-fills and the first London hour. Account equity: 275 USD, including a historical 100 USD welcome credit with wagering conditions still outstanding.
The math block. Wagering conditions on historical 100 USD no-deposit offers typically required 2-5 standard lots traded before any of the bonus became withdrawable; some offers tied it to a profit target plus volume. At 3 lots required and an average half-pip edge per trade — generous — the trader needs to clear roughly 150 pips of net gain across the required volume to see 100 USD convert. On 0.01-lot scalps paying 0.7 pips in friction, each round trip costs 0.07 USD. To cycle 3 standard lots at 0.01 lots per trade requires 300 round trips — that is 21 USD in spread cost alone before any directional thesis.
Hartnett's thesis matters here in a very different way than for Scenario 1. A risk-off tape until DXY peaks means wider ranges in Asia opens, more two-way volatility in EUR/USD during London, and — critically — bigger weekend gaps as positioning adjusts around headlines. For a Sunday-evening gap scalper, that is the environment. More gap. More fill. More opportunity, but also wider stops needed.
Platform-wise, this trader stays on MT4 — not MT5. The reason is specific: MT4's EA ecosystem for scalpers is deeper than MT5's, FBS supports it natively, and the trader's existing indicator stack was built for MT4. Switching to MT5 would gain faster strategy tester and more order types, but lose the custom EA. For scalping where the edge is in trade selection, not order type, MT4 wins. If this trader were running an options overlay or multi-asset, MT5 would win decisively. They are not, so it does not.
The honest read: this trader should not treat the Hartnett note as a signal to add size. The promo-funded structure means the account is operating on borrowed capital with wagering constraints. The macro call simply means the environment favors their style for a few more months. Grind the wagering requirement down. Convert the bonus. Then re-evaluate.
Scenario 3: The Multi-Currency Macro Allocator Sitting on Cash
The third composite is the one that reads Hartnett's note most literally. Imagine an allocator with 50,000 USD split across two accounts — one at AvaTrade for the AvaOptions platform (the only piece of the retail landscape where listed FX options are actually tradeable on a self-directed basis without a prime broker), one at Exness for the 0.1-pip Pro spreads on spot. This is somebody who thinks in portfolio terms, hedges tail risk with options, and uses spot for directional expression. They came to the note already 60% in cash.
The math here is less about leverage and more about scenario-weighted expected value. Let us say the allocator wants to express the Hartnett thesis — risk-off until DXY peaks — as a trade with defined downside. The spot leg: short EUR/USD, 1 standard lot at Exness Pro, 0.1-pip spread. Notional: 100,000 EUR, roughly 109,000 USD. Required margin at 2000:1: 54.50 USD. The option leg at AvaTrade: a 3-month EUR/USD call at a strike 2% above spot, as insurance against being wrong on the top call. Let us say the call costs 0.4% of notional — 436 USD on 100,000 EUR equivalent.
Combined, the position risks 436 USD of premium plus whatever the spot leg bleeds before the stop, in exchange for unlimited downside participation if DXY grinds another 4-6% before topping. Break-even on the structure requires EUR/USD to drop roughly 0.9% from entry — the 0.4% premium plus the spread and swap drag — within the three-month option window. If Hartnett is right about the direction and the timing (dollar peaks sometime inside those three months but after further strength), the structure prints.
If he is right about direction but wrong about timing — the peak comes before the three months are up and the subsequent reversal eats the spot profits — the option expires worthless and the spot leg has to be exited at the trailing stop. Loss is capped at premium plus trail distance.
Platform matters here more than in either prior scenario, and this is where the proprietary-platform-versus-MetaTrader-versus-cTrader argument actually bites. AvaOptions is a proprietary platform. It is not MT4, it is not MT5, it is not cTrader. For listed FX options on a retail account, it is close to the only game in town among the brokers in our verified set. AvaTrade's 400:1 leverage cap is restrictive for pure spot speculation, but for an options-hedged structure that cap is irrelevant — the position is already risk-defined. The allocator uses AvaTrade for AvaOptions and Exness for the spot leg because each platform wins at exactly the thing being asked of it.
This is the deep-detail nerd point we have been building toward. OK, so here is where it gets really interesting: the platform-choice question is almost never answered correctly in retail content because retail content treats platform as a feature list — "MT5 has more indicators than MT4" — rather than as the actual constraint on strategy expression. The platform decides what trades you can structure. AvaOptions is the reason this scenario even exists in a retail context. Without it, listed FX options require a prime brokerage relationship most retail accounts cannot access. The allocator is not choosing AvaTrade for the broker. They are choosing AvaTrade for the platform, and the broker is incidental.
What All Three Share: The Platform Decision Hiding Inside the Macro Call
The three scenarios differ in size, style, and sophistication. What they share is the pattern that gets missed in every discussion of a macro call like Hartnett's: the operational translation of a thesis is a platform decision, not a direction decision.
Scenario 1 needs server-side trailing stops with fine granularity because the trader will not be at the screen when DXY tops. That rules out any platform where trailing is client-side or stepped too coarsely. MT5 at HFM wins. Scenario 2 needs the deepest EA ecosystem for scalping Sunday gaps and London opens — MT4 remains the correct choice, specifically because the strategy's edge sits in indicator stacking, not in order types MT5 would unlock. Scenario 3 needs listed options plus tight-spread spot execution on separate venues, because no single broker in our verified set does both well — AvaTrade for AvaOptions, Exness Pro for 0.1-pip spot.
In each case the macro call is identical. The execution is unrecognizable from one scenario to the next. That is because platform is not a cosmetic choice. It is the layer where the thesis either becomes a tradeable structure or stays a tweet. The retail content industry has trained traders to pick a broker on spread, leverage, and bonus — surface metrics that any affiliate dashboard can rank. The actual choice is which platform lets you express the trades you intend to run. If your style never uses listed options, AvaOptions is irrelevant to you. If it does, AvaOptions is almost the only path. If you scalp with custom EAs, MT4 is correct even in 2026. If you run multi-asset portfolios with options overlays, MT5 is correct. cTrader lives in the middle — superior execution and depth-of-market visualization — but none of the five brokers in our verified set quote on it, which tells you something about where the retail execution ecosystem has consolidated.
Which Scenario Is You: Reading Your Own Setup Against Hartnett's Thesis
Before you add or trim a position in response to Hartnett's call, run a short self-audit. Which of the three scenarios most resembles your account? If you are a carry grinder with directional exposure already on, your question is where the trailing stop lives and whether your platform executes it server-side — not whether to add. If you are a scalper on a small account with any form of promotional overhang, your question is whether the environment favors your style and whether the wagering math still points toward conversion — not whether to double size on a macro note. If you are an allocator with multi-leg structures in mind, your question is which two platforms you need to run in parallel, because no single broker gives you both the options venue and the tightest spot spreads.
The thesis is cheap. The execution translation is the entire trade. Hartnett wrote one note. Three readers should produce three different responses.
FAQ
What exactly did Hartnett argue in the note the article references?
Our reading, grounded in the public summary circulating from Bank of America's weekly flow note, is that Michael Hartnett's desk expects the risk-off regime to persist until the US dollar prints its cycle high, with the reversal in positioning only becoming tradeable from the dollar peak forward. The piece treats the thesis as given and focuses on how three different retail-account structures should translate that view into actual trades, since the macro call is directional, not operational.
Why does the article keep going back to platform choice instead of just naming a trade?
Because platform decides what trades are expressible on a retail account. A carry trader without server-side trailing on their platform cannot hold overnight USD exposure through a Hartnett-style reversal risk without being glued to the terminal. An allocator without access to a listed FX options venue cannot build the risk-defined structure Scenario 3 describes. Platform is not a cosmetic feature list — it is the constraint layer that decides which strategies are even available to run on your account.
Is MT4 really still relevant in 2026 or should everyone move to MT5?
MT4 remains the correct platform for scalpers running custom EA stacks and for traders whose indicator libraries were built before MT5's language shift. The migration tax of rewriting a working EA ecosystem rarely pays for itself unless the trader also wants MT5's multi-asset coverage or the deeper strategy tester. For single-pair FX scalping with mature indicator stacks, MT4 is still active across FBS, Exness, HFM, AvaTrade, and FXTM — the brokers in our verified set all support both.
What is AvaOptions and why does it matter for the third scenario?
AvaOptions is AvaTrade's proprietary platform for listed vanilla FX options, available on self-directed retail accounts without requiring a prime brokerage relationship. In our verified broker set, it is the only platform that lets a retail allocator build spot-plus-option structures natively, which is why the third scenario pairs AvaTrade for the hedge leg with Exness Pro for the tight-spread spot leg. Without AvaOptions, the third scenario's structure would require a different account tier most retail traders cannot access.
How did the no-deposit bonus landscape change and why does it matter for Scenario 2?
Pre-2010 no-deposit offers were marketing experiments with light wagering conditions; the 2018 CySEC restrictions on EU bonus marketing and the parallel ASIC moves around 2020 pushed the model into heavier wagering requirements and non-EU licensing. The historical offers referenced — XM's 30 USD, FBS's 100 USD, Tickmill's 30 USD — operated under those post-restriction rules, which is why the Scenario 2 math shows most of the bonus value getting consumed by spread cost before conversion.
Does the leverage number — 1000:1 at HFM, 2000:1 at Exness, 3000:1 at FBS — actually matter at the sizing described?
At the sizing in Scenarios 1 and 3 — 0.25 lots and 1 standard lot respectively — the headline leverage number barely matters because the effective exposure is small relative to equity. What matters is the headroom it gives for free margin to absorb drawdown without triggering a margin call. For Scenario 2's small-account scalper, the 3000:1 at FBS is more marketing than operational — the trader is sizing 0.01 lots regardless of the cap.
Why did the article mention cTrader if none of the listed brokers quote it?
To make the opposite point. cTrader offers execution and depth-of-market visualization that would benefit the Scenario 1 and Scenario 3 traders if it were available on the brokers in our verified set. Its absence from AvaTrade, Exness, FBS, FXTM, and HFM is itself the data point: the retail execution ecosystem has consolidated around MetaTrader and proprietary platforms, which constrains strategy choice for traders who do not already have an account on a cTrader-quoting broker outside this set.
If Hartnett is wrong about the timing, which scenario is hurt most?
Scenario 3 is most exposed to timing error because the option leg has a defined expiration. If the dollar peak arrives inside the three-month option window but the subsequent reversal is sharp enough to erase the spot leg's gains before expiry, the structure finishes near break-even or at a loss capped by premium and trail distance. Scenarios 1 and 2 are timing-insensitive — the carry grinder's trailing stop manages exit regardless of when the top prints, and the scalper's edge is intraday and does not depend on the macro call resolving.