For Corporate Treasury & CFOs

A hedge with a defined edge, built on your cycles — repeatable, systematized, defensible.

CycleHedge maps three cycles — your business exposure, the currency pair's measurable rhythm, and the interest-rate differential driving forward points — into weekly hedge signals. Every decision timestamped, every signal logged. Nine years out-of-sample. Every year an edge over fifty-fifty.

Schedule Your CycleHedge DemoSee the 9-year proof

Somebody on the oversight committee — the board, the audit committee, the risk management committee — is going to ask the question differently this quarter, but the wording changes and the question does not.

“Walk us through your FX risk management framework.”

The honest answer, for most treasuries, is that there is no framework. There is a policy, a single rule inherited years ago. There is a bank email, which is a view.

There is a spreadsheet, which is a record of the hundred-percent forward cover that always gets placed because you got badly burnt when you were open and the market moved against you — and now don't want to be at risk.

Or the opposite— a hundred-percent Spot policy — for exactly the same reason (you got badly burnt when you took cover and the market moved in your favour, creating a huge forex loss), and for this reason you don't want to be at risk.

Or you have tried both, and they have both worked for a while, until they didn't— so now you sit on the fence, taking fifty-percent cover upfront and leaving fifty-percent uncovered, thinking this is the best you can do, but knowing there must be a better way.

That is the specific moment the page you are reading was built around.

Bring your company's parameters. See your nine-year backtest. Decide with evidence.

Schedule Your CycleHedge Demo

You are trying to solve it as a prediction problem. But what you have is an FX risk management problem — and governance is what will expose that.

Every tool you have bought for your treasury was built to help you predict currencies better — bank advisory desks, Bloomberg terminals, treasury management platforms, macro research subscriptions. Your entire vendor stack sits on one side of a divide most treasurers have never had named out loud.

The divide is this.

A prediction system is evaluated by whether its forecast came true.

An FX risk management system is evaluated by whether its methodology is sound. That evaluation is what your risk, audit, and board committees are there to do.

For context on why we are the ones drawing this distinction: the analyst team behind CycleHedge publishes a twenty-one-year prediction record — 9,690 forecasts, 8,756 scored against actual outcomes across currency pairs, 72.3 percent directional accuracy, all auditable. We hold the prediction credentials. What two decades of running them has made clear is that prediction credentials are not the document the committee is asking for.

The committee reviewing your hedging program — whether it sits as the board, the audit committee, or a standalone risk committee — is not running an accuracy check on your banker's six-month FX call. They are asking whether the process that produced your hedge placements was systematic, documented, and defensible in the same way a prudent-person review would defend any other fiduciary decision. Their job is the evaluation. Your job is to give them a methodology worth evaluating.

That is a different kind of question. And it is the one the industry has never built a product for.

Which is why your tools have improved every year and your anxiety before board meetings has not.

Why information alone is not enough.

Even when the information is correct, even when the timing is right, the decision still lands on a human. And the human is part of the crowd. The human is second-guessing. The human is the one whose name will be on the hedge placement if it moves against them.

Over twenty-one years I have fielded phone calls from clients who had the forecast in front of them, published, timestamped, unambiguous, asking whether they should act now or wait until tomorrow. They had the information. They had the timing. What they did not have was protection from their own emotions at the moment of decision.

Our published forecasts still carry a subjective layer. They tell you where the cycle is and where the market is likely to go. What you do with that, and when, is still your call.

CycleHedge removes that layer. The signal tells you what to do and when to do it. The committee reviews the log. The CFO defends the system, not the call.

I have had this conversation with more than one CFO operating at scale. Their point, in their own words, is that they need something to defend to the growers, boards, and committees they answer to. A system with a documented edge gives them a fall-back when things do not go to plan; a personal judgement call does not. Accountability sits with them either way, but one version gives them an audit trail to stand on, and the other leaves them standing alone.

The Indefensible Blank

The problem is not the question. The problem is what sits behind it.

If you have sat through the question, you know the problem is not the answer. The problem is that there is nothing systematic behind the answer.

You have spreadsheets. You have a bank email. You have a policy that says “always hedge fifty percent rolling, unless the CFO decides otherwise.” None of that is a framework. It is policy, opinion, and post-hoc narrative. It works in the room until somebody competent asks a second question.

And the second question, “what would have changed your reasoning?”, is the one that ends mandates.

Not because the hedge lost money. Boards do not fire treasurers for FX losses. They fire treasurers who cannot explain the process that produced the loss.

The stake is not the exchange rate. The stake is the methodology you can present when the market moves against you.

Five patterns that do not survive the audit.

Every treasury I have looked at runs one of these patterns. The first three are structured policies, each with a trauma behind it. The last two are what the treasury falls back to when the structured policy has already failed somebody in the room.

The Hundred-Percent Cover.

A policy of always hedging the full exposure, rolling — put in place by somebody who got badly burnt when the book was open and the market moved against them. It locks in a bad rate every time the cycle is moving in your favour, and it locks out the upside when the cycle turns. The decision was made once, under duress, by someone who is probably no longer in the seat. It cannot be defended as methodology because it was never methodology; it was a scar.

The Hundred-Percent Spot.

A policy of never hedging — take the Spot at the moment of receipt. Put in place by somebody who got badly burnt when cover was taken and the market moved in the company's favour, creating a forex loss on the hedge. The reaction is the mirror of the first. The consequence is the mirror of the first: full exposure to the adverse side of every cycle. Same trauma, opposite reaction, same absence of system.

The Fifty-Fifty Fence.

After trying both — and finding that both worked, until they didn't— the treasury sits on the fence. Fifty-percent cover rolling, fifty-percent uncovered, on the theory that it splits the difference. It does not split the difference; it guarantees mediocre outcomes across every cycle. When the board asks “why fifty?” the only honest answer is “because we tried everything else.” That is not methodology. That is exhaustion.

On top of all three of these policies — and often instead of them, once the structured policy has stopped being defensible — sits the less structured approach. When the policy is no longer carrying the argument in the room, the treasury falls back to something that at least feels like a judgement. It is not a methodology either. It is just a different kind of not-methodology.

The Judgement Call.

Someone senior “reads the market” and times the cover. When it works, nobody asks. When it does not, the board wants to know why the company locked in at the worst rate of the quarter. There is no defensible answer if the decision was based on feel.

The Bank's View.

Your relationship manager has a call. You know, privately, that the bank earns the spread whether the call is right or wrong. You also know, because every treasurer who has been in the seat for more than five years knows, that consensus bank FX forecasts have been documented to lose to a coin flip at most horizons (Meese & Rogoff, 1983; confirmed in subsequent studies). You quote the bank anyway because it is the only narrative available.

All five share one structural failure. The decision has no system behind it. There is nothing to audit. Nothing to defend. Nothing to show the committee beyond “this is what we did, and we thought it was the right call at the time.”

That is not a system. That is the absence of one.

Yet you know there must be a better way. After two decades watching every variant of these five patterns fail the same audit question, we built what was missing — a system with a measurable, out-of-sample edge over each of them, tailored to your exposure, your currency pair, and the phase of the cycle your book actually runs on.

Three cycles — one defined edge, built for your exposure pattern.

Three kinds of cycle run through every corporate FX book, and no vendor has ever connected them into a single timing discipline. CycleHedge reads all three and converts them into a weekly hedge signal.

1

The operational cycle.

The rhythm your book runs on — your financial year (or your season, if you are in agriculture) and the exposure windows inside it: receivables, creditor payments, the forward horizon you actually hedge to. The point at which exposure gets realised, and the decision window in which you can act on it.

2

The market-structural cycle.

Currency pairs move in measurable, recurring patterns. Not random. Not perfectly periodic. Cyclical in the specific sense that the phase of the cycle is identifiable in real time from the data, and the historical record shows that decisions made with reference to the phase have measurably outperformed decisions made without it. Point forecasts lose to a coin flip at most horizons. Position-sizing decisions taken against phase-of-cycle do not, because the decision is probabilistic, not directional, and the value is harvested across the sequence, not extracted from a single call.

3

The rate-structural cycle.

The rate differential between the two currencies in any pair sets the forward-point curve. Where the higher-rate currency is the one you pay (you import priced in it), holding forward costs you every day you hold. Where the higher-rate currency is the one you receive (you export priced in it), holding forward pays you. The differential determines who benefits from timing, and how much timing is worth in your reporting currency per year. The economics are universal; the magnitude varies by pair — most pronounced on emerging-market pairs like USD/MXN and USD/ZAR, measurable on USD/CAD, EUR/USD, and the G10 majors.

Rates and signals — the timing signal. When the forward is worth taking, and when the future spot is. Read weekly, logged every time.

CycleHedge integrates the three into a weekly HEDGE / WAIT / STOPPED decision on the cover ratio of your book. Every decision is timestamped, every decision logs its inputs, and the committee reads the ledger by week, by pair, by period — the call stops being yours to defend.

The oversight committee — board, audit, or risk — does not need to trust your judgement. They can read the ledger.

This is the defined edge, built on your cycles, that the prediction tools never delivered — repeatable every exposure period, systematized every week, defensible every time the committee asks.

The Proof

Every call the mechanism makes is written down. Every signal, every trigger level, every committee sign-off, every placement versus the 50/50 benchmark to the basis point. The audit trail is how the committee reviews what the mechanism did — and it is how we can show you nine years of performance that is verifiable week by week.

The audit trail. One row per week — pair, direction, phase read, trigger level, committee sign-off, and placement versus the 50/50 benchmark to the basis point. Every decision timestamped. Every override logged. The document the committee evaluates.
Schedule Your CycleHedge Demo

Nine years out-of-sample, with every year positive.

The methodology was parameter-set on 2014-to-2016 data. Performance was measured on 2017-to-2025, 467 weeks the model never saw during design. Against a fifty-fifty benchmark (fifty percent forward cover, fifty percent spot at receipt, the industry-standard reactive approach), the CycleHedge timing layer delivered a documented edge every single year. Including 2020. Including 2022. Worst-month drawdowns were recovered within three to four weeks.

Cumulative basis-point edge versus the 50/50 benchmark, plus annual performance by calendar year. Nine-year out-of-sample record, every year positive.

Applied to a trading book that ranged between R1.3bn and R1.7bn (approximately $75M to $95M USD) across the nine-year period, averaging about R1.65bn (~$90M), a 0.71% average edge over the 50/50 benchmark produced R11.7 million per year on average (~$650k), accumulating to R105.7 million (~$6 million USD) with every one of the nine years positive, including COVID. For a book of that scale, that is roughly three full years of a Finance Director's total compensation package — recovered from timing decisions that never required a capex approval.

Edge is measured as the realised rate (in your reporting currency) on the covered portion versus the 50/50 reactive benchmark across the same weekly decision points. Worst-year edge, best-year edge, and the full distribution by year are walked through in the demo on your own parameters.

CASE STUDY: A live three-year engagement at the same scale.

Before the productised rebuild, the same methodology ran as a human-in-the-loop advisory engagement with the largest citrus exporter in the Sundays River Valley from 2017 through 2019. The engagement hedged the export book's three major currency exposures — USD/ZAR, EUR/ZAR, and GBP/ZAR — on a rolling ten-week forward exposure window, week by week, across three consecutive years. Three consecutive positive years versus the fifty-fifty benchmark. In 2019 alone, on a book of R1.49 billion (~$100M USD at 2019 rates), the timing value attributed to the engagement was R35 million-plus (~$2.4M+). The three-year average uplift was 1.19 percent.

The live engagement uplift (1.19%) runs higher than the nine-year backtest edge (0.71%) for two reasons, both declared openly. First, 2017-to-2019 was a regime favourable to the methodology across all three pairs the engagement hedged, which the full nine-year out-of-sample record dilutes rather than hides. Second, the human-in-the-loop layer added judgement on the margin that the automated product deliberately does not replicate, because the repeatability is the point. CycleHedge is built to hold the 0.71% edge across every regime, not to chase the 1.19% peak of one.

The engagement was structured around a Risk Management Committee: every decision went through committee review, every weekly signal was documented, and every quarter's performance was audited against the benchmark. That structure was not bolted on — it was the point, and it is the same structure CycleHedge now runs automatically. The client's CFO takes reference calls on request, thirty minutes, audited methodology questions, no sales presence. We do not publish written testimonials, because this kind of work earns its proof in a private conversation between peers, not in a quote tile on a page.

The analyst team's pedigree, not the CycleHedge proof.

The cycle-timing discipline underneath CycleHedge is the same one underneath the published Dynamic Outcomes forecast record: twenty-one years, 9,690 published forecasts, 8,756 scored against actual outcomes, directional accuracy of 72.3 percent, all published, all auditable. The forecast record is the pedigree of the analyst team. It is not the CycleHedge proof.

CycleHedge does not require directional accuracy on any single forecast to work. It requires consistent phase-of-cycle identification, which is what the nine-year out-of-sample record above measures. The public forecasts and CycleHedge answer different questions. The forecasts tell you where currency pairs are likely to go. CycleHedge tells you what to do and when to do it.

The methodology was originally built on USD/ZAR — the pair with the longest clean backtest horizon, the deepest rate differential, and the most pronounced cycle amplitude. The same three-cycle discipline was live-engagement validatedon EUR/ZAR and GBP/ZAR in the 2017–2019 advisory engagement above, and has since been framework-validated across EUR/USD, USD/CAD, and USD/MXN. The principle does not depend on which currency sits on which side of the pair. It depends on whether the book has forward-hedgeable exposure and a committee that audits the decision.

You do not need to take any of this on trust. The nine-year backtest runs in a demo on your pair, your exposure direction, your decision window. If the edge is there for your specific profile, you see it in the ledger. If it is not, you see that too. The demo is the decision tool.

Run the Backtest on Your Parameters
The Cost of Not Having It

Two treasuries running the same book across the same cycles — on different systems.

Consider two treasuries that started this year with the same book, the same currency exposure, the same quarterly board rhythm.

The first continues the approach it has always used. Fifty-fifty cover rolling, with occasional overrides when the banker's view is confident enough. Over the next nine years, across currency cycles that will include at least one COVID-scale dislocation and at least one rate-differential-driven currency repricing, the treasury will produce results that, looking at the data rather than the narrative, average out to roughly what a coin flip would have produced. On a R1 billion (~$55M) book, the cumulative cost to the firm sits in the range of R30 to R100 million (~$1.7M to $5.5M) in timing value left on the table.

The second treasury adopts CycleHedge — a hedge with a defined edge, built on its own cycles. Same markets, same volatility, the same cycles of currency strength and weakness — different system. Every board meeting, the treasurer walks in with a documented methodology, a weekly signal log, and a nine-year audited track record of the approach. The market moves against a position? Fine — the framework explains the position, and the board sees institutional excellence, not guesswork.

At the scale documented above, the difference between the two treasuries over nine years is R105 million (~$6M USD) in the firm's favour, and a replacement for the methodology gap that sits underneath every board conversation.

The question is not whether the edge exists. The backtest has already answered that. The question is whether you want to be the treasury that has it.

Schedule Your CycleHedge Demo
Who This Is For

Corporate treasury teams who answer to the board and need a documented methodology.

Corporate treasuries and CFOs running annual FX exposure at the mid-market-to-enterprise scale, with board or audit-committee oversight of hedging decisions, who need documented, defensible timing methodology — not another prediction tool.

Currency pairs

  • USD/ZAR — deep validation (the pair the nine-year backtest was run on, and one of three pairs in the live three-year engagement).
  • EUR/ZAR, GBP/ZAR — live-engagement validation (the other two pairs in the 2017–2019 advisory engagement).
  • EUR/USD, USD/CAD, USD/MXN — framework validated across G10 majors and USD-emerging crosses.

Scale coverage

Validated up to R1.7bn (~$95M USD) book scale. The methodology is notional-agnostic; execution considerations above R2bn (~$110M USD) — forward-curve liquidity, placement sizing relative to dealer appetite — are discussed in the demo and sized to your counterparty stack.

Where CycleHedge fits

Bank-agnostic — your existing counterparties execute and your existing treasury platforms run. CycleHedge is the timing-intelligence layer, and it replaces the judgement call, not the bank relationship.

Where it is not a fit

Treasuries that prefer a static, never-review hedging policy; companies where board and audit oversight is informal; books where the timing value does not clear the cost of implementation.

In Practice

How it lands in your week.

Monday morning — the weekly signal drop arrives by email, delivered before European markets open: pair, direction, phase read, execution window, trigger level. Same format every week, same format every year.

Your treasury analyst (or you) executes through your existing bank counterparties, and the audit log generates on the fly — CSV export to your TMS, PDF board pack auto-built on month-end.

Onboarding runs two to four weeks including one cycle of parallel-signal review. When the signal disagrees with judgement, your internal override protocol logs both the signal and the override with a signed-off reason.

The system makes the call. The record survives the audit.

Engagement

Priced as a fraction of the documented saving.

Annual engagement scaled to book size. Starting at $15,000 for smaller treasuries and tiered on basis points of book thereafter, priced as a fraction of the documented saving.

Structured as annual retainer, not per-transaction. Bank-agnostic. No commission. No spread participation. The demo calculates your specific number before any commercial discussion.

The Crossroads

You are a decision away from the answer.

You are not a year away from answering the question your oversight committee is going to ask — you are a decision away, and right now you are standing between two paths.

One path continues the policy you inherited: the fifty-percent rolling cover that came from nobody remembering who chose it, the bank email, the spreadsheet, the post-hoc narrative you have rewritten so many times you no longer hear yourself doing it.

The other path is the framework you can hand to the committee — board, audit, or risk — the same way you hand them the annual audit file. Nine years of out-of-sample evidence on a R1.3-to-R1.7-billion (~$75-95M USD) book scale. A weekly signal log on your pair, your direction, your decision window. A mechanism that reads the three cycles and places the cover where the data says the cover belongs, with every decision timestamped and every decision auditable.

I have watched every variant of the five patterns fail, and I have watched one methodology, this one, deliver a documented edge every single year against the benchmark, including the years the market nearly broke every treasury in the room. I am not asking you to trust that. I am asking you to run it on your numbers.

Schedule a demo. Bring your pair, your exposure direction, and your decision window. We will run the nine-year backtest against your parameters. You will see the edge for your profile, or you will see that it is not there. Either way, you will know. Twenty minutes. The meeting is the decision tool. No commitment to anything beyond the meeting itself.

The data decides.

Schedule Your CycleHedge Demo
James Paynter

James Paynter

Principal, Dynamic Outcomes

Twenty-one-year published currency-cycle track record

Direct line for reference discussions: james@dynamicoutcomes.io

CycleHedge is a hedge timing decision-support tool, not financial advice. The nine-year out-of-sample backtest (2017–2025) and the documented three-year live engagement (2017–2019) reflect specific test and engagement conditions; past performance does not guarantee future results. All hedging decisions remain the responsibility of the subscribing organisation. Dynamic Outcomes (DBA Dynamic Forex Solutions) — 21 years of published FX forecasts — 9,690 documented forecasts, 8,756 scored against actual outcomes, directional accuracy of 72.3%.