This is investment research, not personal financial advice.

IDP Education (ASX:IEL) closed at A$1.72 on Thursday, down 20.7 per cent, after filing its FY26 results at 8.09am. About 30.8 million shares traded, 6.4 times the twenty-day average, and A$125 million of market value left the register in a single session (ASX 2026). Reuters called it a record low; on the numbers this desk checked, it is at minimum the lowest close in the past year, beneath a prior floor of A$2.00 (Reuters 2026). The strange part is what the release actually said. The headline read "FY26 results resilient and in line with guidance", and that was true: adjusted EBIT of A$122.9 million landed inside the A$120-130 million range set in February (IDP 2026a). The market read past the headline to the FY27 paragraph, which plans for market volumes to fall another 20 to 30 per cent and guides adjusted EBIT down to A$95-115 million. A third consecutive year of falling profit, with a fourth priced in, is what cost the company a fifth of its value.

The question this article answers is whether that repricing is correct. The short version: on the guidance arithmetic alone, the fall looks close to proportionate. But the accounts also show a company where the better business got more profitable on the day the market flattened both halves of it.

What IDP actually is, and the deal that made it

IDP runs three businesses that most summaries blur into one. The oldest is student placement: IDP counsels international students and places them with universities in Australia, the UK, the US, Canada, New Zealand, Ireland and, from this year, Malaysia and the UAE, earning a commission from the institution on enrolment. The second is English language testing, where IDP is one of three co-owners of IELTS alongside the British Council and Cambridge, and operates the test across most of the world. The third, small and quiet, is English language teaching through schools in Cambodia and Taiwan (IDP 2026b).

The placement agency and the test have different economics, and the difference matters for everything that follows. Placement is an agency business: 82 per cent gross margin on placements in FY26, revenue booked at census date when the enrolment becomes final, volumes set by how many students destination governments let in. Testing is closer to a royalty on global mobility: a 13-day service window, fees around A$325 a test, and a recognition network IDP counts at more than 13,500 organisations, which is the asset competitors cannot quickly copy (IDP 2026b, IDP 2026a).

The company that listed in 2015 at A$1.55 was mostly the agency. The company that exists today was substantially assembled in August 2021, when IDP paid about A$260 million for the British Council's India IELTS operations and took over the largest English test market in the world's largest student source country (IDP 2022). That acquisition, plus the Intake Education purchase in FY23 for A$81 million, built the FY23 earnings peak. It also concentrated the group in exactly the geography now causing pain: India testing revenue fell 24 per cent in FY26, on regulatory change and, in IDP's own words, "increased competition" (IDP 2026b).

The FY26 accounts: statutory truth against adjusted comfort

The gap between IDP's adjusted and statutory numbers is now the whole story, so it is worth laying out both. On the adjusted basis IDP manages to, revenue fell 9 per cent in constant currency to A$795.4 million and adjusted EBIT fell 7 per cent to A$122.9 million (IDP 2026a). On the audited statements, revenue fell 11 per cent, EBIT fell 38 per cent to A$61.8 million, and net profit fell 74 per cent to A$12.3 million (IDP 2026b). Basic earnings per share were 4.41 cents against 36.86 cents four years ago.

The A$61 million bridge between those two pictures contains five items, and most of them are cash (IDP 2026b, IDP 2026c). Restructuring and transformation costs of A$36.2 million paid for the removal of about 1,250 roles, roughly 20 per cent of headcount, and the shrinkage of the IELTS venue network from about 1,500 sites to fewer than 600. A A$6.0 million credit loss provision covers students in countries subject to foreign exchange controls, the same line that cost A$15.1 million a year earlier. Amortisation of acquired intangibles, including A$6.2 million a year from the Speak technology purchase in March 2025, adds A$10.8 million. Customer remediation and M&A costs make up the rest. A further A$40 million of one-off transformation costs is guided for FY27.

The five-year shape of the business explains why the market has lost patience with adjusted numbers. Revenue is back to exactly its FY22 level; profit is not.

FY Revenue A$m EBIT A$m NPAT A$m EPS c OCF A$m Capex A$m Net debt A$m ROIC % (computed)
2022 793.3 158.9 102.6 36.9 140.6 29.6 (40.2) 26.2
2023 981.9 220.7 148.5 53.4 170.3 38.9 42.4 27.2
2024 1,037.2 210.3 132.7 47.7 116.0 54.2 170.0 21.1
2025 882.2 92.3 44.5 16.0 132.6 60.9 165.1 9.2
2026 795.4 61.8 12.3 4.4 135.5 32.3 119.6 8.3

Figures are as reported in each year's audited Appendix 4E (IDP 2022, IDP 2023, IDP 2024, IDP 2025b, IDP 2026b). Two columns carry this desk's own work and are labelled for it: the ROIC column is an author-computed return metric, statutory EBIT taxed at 30 per cent over net debt plus equity, and the FY22 and FY23 net debt figures are author-computed from the balance-sheet lines, where the later years use the company's stated definition. On that computed ROIC basis, IDP went from a 27 per cent compounder in FY22-23 to 8 per cent in three years. Even on adjusted EBIT, the FY26 figure is about 17 per cent. The reinvestment economics did not gradually soften; they stepped down twice, once with the FY24 volume peak passing and again when the policy caps landed in FY25.

Then there is tax, which deserves more attention than it gets. IDP paid A$27.3 million of income tax on A$40.6 million of pre-tax profit, an effective rate of 67 per cent. The reconciliation shows why: A$15.7 million of unrecoverable withholding taxes, four times last year's A$3.6 million, plus attributed foreign income taxed in Australia without matching credits (IDP 2026b). An English test administered in India, priced in rupees, against intellectual property held in Australia, leaks tax at every border. The final dividend being unfranked tells the same story from the franking-account side: Australian tax paid is shrinking while offshore withholding grows.

One more disclosure belongs in this section because it is the most honest sentence IDP has published in two years. On 18 December 2025 the company voluntarily changed its placement revenue recognition from the point of enrolment confirmation in Australia and the UK to census date, writing A$68.8 million off retained earnings to restate FY25 (IDP 2025a). The stated reason was that "considerations related to Student Placement performance obligations have evolved", and that students now need more support to actually arrive and commence. In plain language: revenue booked when a visa was granted had stopped being a reliable proxy for revenue earned, because too many confirmed students were not converting into enrolled ones. A policy-change announcement is an unusual place to find a confession of demand-quality deterioration, and the market treated it as an accounting item at the time.

Where the volumes went, market by market

The FY26 placement numbers show a collapse that is broad but not uniform. Total placement volumes fell 27 per cent to 54,200. By destination: Canada down 80 per cent, the United States down 66 per cent, the UK down 20 per cent, Australia down 9 per cent, with New Zealand and Ireland flat to slightly up (IDP 2026b). By source: India down 39 per cent, China down 22 per cent, Vietnam down 19 per cent, against Bangladesh up 13 per cent. Average fee rose 10 per cent to A$5,586, which is how a 27 per cent volume fall became only a 17 per cent revenue fall.

This is the second year of it. IDP placed 98,900 students in FY24, 70,100 in FY25 and 54,200 in FY26. Testing volumes peaked at 1.93 million in FY23 and ran 1.20 million in FY26, down 38 per cent from that peak (IDP 2022, IDP 2024, IDP 2025b, IDP 2026b).

The government data confirms the policy, not IDP, is doing the cutting. Australia's Department of Education counts 216,884 commencements in the year to May 2026, down 8 per cent on an already reduced prior year, which had been running 15 per cent below the year before. Enrolments by continuing students are only down 2 to 8 per cent depending on the measure, which is why universities have not yet felt the full effect: the system is losing new entrants fastest, and new entrants are precisely what a placement agent sells (DoE 2026). English language pathway enrolments, ELICOS, are down 27 per cent year to date, the single worst sector in the series and the closest proxy for IDP's testing and teaching funnel. The FY27 guidance assumes this keeps going: IDP is planning for total market volumes across its six destination countries to fall 20 to 30 per cent again, "based on no further change in key immigration and visa policy settings" (IDP 2026a). The phrase is doing heavy lifting. It means the caps already legislated and announced are enough, on their own, to cut the addressable market by another quarter.

Two engines, one valuation

Here is the fact the share price ignored. English language testing gross profit rose 1 per cent in FY26 to A$164.1 million while testing volumes fell 8 per cent, lifting the segment gross margin from 39.5 to 42.2 per cent. Average yield rose 7 per cent in constant currency. The margin gain came from venue rationalisation, the shift to third-party operating models in some countries, a favourable currency effect on the fee IDP pays to Cambridge, and operating leverage in the business that held up (IDP 2026b, IDP 2026c). On the same day the group was marked down 21 per cent, its most valuable franchise improved its unit economics.

Placement also defended its margin, differently: gross profit fell 16 per cent on 17 per cent lower revenue, holding an 82 per cent placements margin, as average fees rose and direct costs fell 8 per cent (IDP 2026b). Yield over volume is working as a strategy in the narrow sense. Whether it can keep working is a fair question, since a 10 per cent fee rise in a market shrinking 27 per cent implies universities still value quality intermediation enough to pay for it, and there is a limit to that at any given volume.

The counter-evidence on the moat is real, and it is concentrated in India. India testing revenue fell 24 per cent, which IDP attributes to regulatory and immigration changes plus increased competition; India is described elsewhere in the report as a higher gross margin country, so share movement there bites harder than headline volumes suggest (IDP 2026b). The competitive set is worth naming with numbers. Pearson's PTE is the aggressive challenger in India on price and result speed. Duolingo's DET, the frequently cited disruptor of high-stakes English testing, reported revenue of US$10.1 million in the June quarter of 2026 against US$10.1 million a year earlier, flat, on a company total of US$298 million (Duolingo 2026). DET at roughly US$85 million a year running rate against IELTS's A$389 million inside IDP alone, with DET growth stalled, reframes the disruption story: the near-term threat in India is a rival incumbent, not a Silicon Valley app.

The durable assets on the testing side did visibly strengthen. IDP renewed the IELTS service agreement with Canada's immigration department long term, opened 13 IELTS centres in China in a phased market entry, and cut the venue network's fixed cost base by two thirds (IDP 2026c). A test accepted by 13,500 institutions and embedded in three governments' migration systems does not lose its franchise to a bad volume year. What it can lose, slowly, is India.

Cash: a 139 per cent conversion rate with an expiry date

IDP's balance sheet is the strongest part of the investment case, and the part where the reported flattery is easiest to miss. Operating cash flow was A$135.5 million, up on the prior year despite statutory profit collapsing to A$12.3 million, and cash conversion printed at 139 per cent of EBITDA (IDP 2026b, IDP 2026c). Three tailwinds made that number. Working capital released A$35.8 million, as receivables fell 41 per cent and contract assets 28 per cent on lower volumes and faster billing, a release that cannot repeat once those balances finish unwinding. Income tax paid fell from A$65.0 million to A$16.7 million, a A$48 million swing that reverses as the withholding position normalises. And capex was cut 47 per cent to A$32.3 million, with the deferral of technology projects partly offsetting it. Strip the working-capital release and normalise tax toward A$30-odd million and underlying operating cash is closer to A$80-90 million than A$135 million. Less capex, that supports an owner cash flow of roughly A$50-60 million before the A$36-40 million a year of cash restructuring, or A$85-95 million if restructuring is treated as one-off. Both figures are computed by this desk from the filed statements, not reported by IDP.

None of this makes the balance sheet fragile. Net debt ended at A$118.6 million, the banking covenant leverage ratio at 1.0x against 1.4x, with A$341 million undrawn across A$595 million of unsecured facilities maturing between 2028 and 2030 (IDP 2026b). Six months earlier, at December 2025, net leverage was 2.5x, so the second half deleveraging was fast and real (IDP 2026f).

Capital allocation now runs through three channels. The dividend was cut to 9 cents for the year, 3 cents franked at 50 per cent and a 6 cent unfranked final, down from 14 cents and 34 cents two years before that; the payout on adjusted earnings is under a third, which is prudent for a business guiding lower (IDP 2026a). The A$50 million on-market buyback announced on 19 June had barely started by 31 July, when a broker was formally appointed, suggesting the program was paused for the results blackout; at A$1.72 the full amount retires about 10 per cent of the register (IDP 2026e). And capex discipline, A$32 million against A$61 million, is a choice about how much of the AI and ERP agenda gets funded from internal cash. A board buying back stock with one hand while cutting reinvestment with the other is making a statement about where value lies; the statement can be right or wrong, but it is coherent.

What the fall actually repriced

Some arithmetic on the repricing. Before the result, at A$2.17, IDP's market value was A$604 million and enterprise value about A$723 million, or 5.9 times FY26 adjusted EBIT and 6.9 times the midpoint of FY27 guidance. After, at A$1.72, enterprise value is A$597 million: 4.9 times FY26 adjusted EBIT, 5.7 times the FY27 midpoint of A$105 million, and about 8.4 times FY26 adjusted NPAT of A$57.1 million (all multiples computed by this desk from the filed figures).

Model FY27 at the guidance midpoint and the same conversion IDP achieved in FY26, with net finance costs around A$19-21 million and an effective tax burden in the low thirties plus irrecoverable withholding, and adjusted NPAT comes out around A$55-58 million, flat to slightly down on FY26. In other words, A$1.72 is the FY27 guidance midpoint carried at roughly the trailing multiple. The pre-result price needed either the top of guidance or a re-rate to justify itself. The market chose neither, and the 20.7 per cent fall decomposes into about a 15 per cent downgrade of the earnings base plus a further markdown of the multiple for terminal growth. That is the shape of a market deciding the decline does not end in FY27.

The four scenarios below are built on adjusted NPAT times a multiple, with the FY27 guidance range and the policy data as the swing variables. The severe case has India competition and a worse-than-guided volume outcome cutting adjusted EBIT to A$80 million and NPAT to A$32 million at 7 times. The bear case delivers the guidance floor with FY28 still falling. The base case delivers the midpoint with volumes stabilising at the administered floor, which is not a forecast of good news, only of no further bad news. The bull case needs policy easing in at least two of the six destination markets plus India share stabilisation, which lands FY28 adjusted NPAT near A$70 million at 11 times.

  • Severe downside: A$0.80-1.05
  • Bear: A$1.10-1.40
  • Base: A$1.65-1.95
  • Bull: A$2.40-3.10

At A$1.72 the stock trades at the low end of the base case. The sensitivity that matters most is not the multiple, it is the FY27 EBIT line: every A$10 million of adjusted EBIT near the midpoint is worth roughly 16 cents a share at a constant 9 times NPAT conversion, so the gap between a floor scenario and a continued-contraction scenario spans the entire band between the bear and bull cases.

Run the valuation in reverse and today's price says this: IDP generates A$50-60 million of owner cash, that number does not grow, and an 11 per cent cost of equity on a stock with 64 per cent 90-day volatility is fair compensation. Thursday's close underwrites either the policy floor or the buyback, not a recovery.

The crux, the calendar, and what answers it

Three facts decide whether the bear or base case is right, and each has a date.

First, guidance credibility. H1 FY26 delivered A$87.5 million of the full year's A$122.9 million adjusted EBIT, 71 per cent, and management has flagged FY27 is weighted the same way while the next A$15 million of cost-out lands mostly in the second half (IDP 2026f, IDP 2026a). The half-year report around February 2027 that prints below roughly A$65 million adjusted EBIT puts the A$95-115 million range under visible pressure. The prior year's upgrade cycle shows the mechanism works in both directions.

Second, the policy floor. The Department of Education's monthly commencement series is the highest-frequency instrument anyone has, and it currently reads minus 8 per cent year to date after a prior year that was down 15 per cent (DoE 2026). The FY27 plan needs that series to stop deteriorating without any further policy tightening. Australian settings land in the federal budget cycle; Canadian and UK policy moves arrive on their own timetables. A stable administered quota is a strange kind of good news, but it is the specific good news this valuation requires.

Third, India. A second consecutive year of double-digit decline in India testing revenue converts the story from market shrink, which is cyclical-policy, to share loss, which is structural. The half-year service segment table answers it in February.

Beyond those three, the monitoring list this desk will run: commencement data monthly; the Appendix 3Y buyback notices for evidence the board still wants the stock below A$2; net leverage against the 1.5x commitment through FY27; and the pace of the China IELTS rollout, the one new revenue stream in the plan.

Two things are missing from this research and are said so plainly. The FY27 guidance's volume assumption is derived from destination-market policy announcements this desk has not independently totalled, so the 20-30 per cent figure is IDP's, not ours. And the record-low characterisation of the share price is Reuters', taken from its headline; this desk verified the 52-week-low claim from its own price data and left the all-time claim attributed.

The verdict on the reaction, for the record: close to proportionate, with a small overshoot at the close. A 15 per cent earnings-base downgrade at the previous multiple accounts for most of the fall; the remainder is a repricing of what comes after FY27. What the market is now pricing is a company that earns A$55-60 million adjusted, pays a third of it out, buys back its own stock at a tenth of the register, and stops shrinking. The accounts filed on Thursday show half of that already being true. The half that is not, the part where the test franchise keeps compounding while the agency keeps cutting, is exactly what the monthly commencement series and the February half-year will confirm or refute.

References

  • ASX 2026, ASX company page for IDP Education Limited (IEL), 20 August 2026.
  • IDP 2026a, IDP Education Limited, FY26 results release, 20 August 2026.
  • IDP 2026b, IDP Education Limited, 2026 Appendix 4E and Annual Report, 20 August 2026.
  • IDP 2026c, IDP Education Limited, FY26 Financial Results presentation, 20 August 2026.
  • IDP 2026d, IDP Education Limited, Appendix 3A.1 final dividend notice, 20 August 2026.
  • IDP 2026e, IDP Education Limited, Appendix 3C buy-back update, 31 July 2026.
  • IDP 2026f, IDP Education Limited, H1 FY26 results release, 26 February 2026.
  • IDP 2025a, IDP Education Limited, revenue recognition policy change announcement, 18 December 2025.
  • IDP 2025b, IDP Education Limited, 2025 Appendix 4E and Annual Report, 27 August 2025.
  • IDP 2024, IDP Education Limited, 2024 Appendix 4E and Annual Report, 28 August 2024.
  • IDP 2023, IDP Education Limited, 2023 Appendix 4E and Annual Report, 22 August 2023.
  • IDP 2022, IDP Education Limited, 2022 Appendix 4E and Annual Report, 24 August 2022.
  • DoE 2026, Australian Department of Education, International student monthly summary, YTD May 2026.
  • Duolingo 2026, Duolingo, Inc., Form 10-Q for the quarter ended 30 June 2026.
  • Motley Fool 2026, The Motley Fool Australia, IDP Education posts steep FY26 profit drop but stays on transformation track, 20 August 2026.
  • Reuters 2026, Reuters, IELTS exam co-owner IDP Education shares hit record low after flagging weaker 2027 earnings, 20 August 2026.

Source notes. All financial history in the table and text is from IDP's audited Appendix 4E filings for each year, as originally reported for FY22-FY25 and on the restated basis where the FY26 report states one; the December 2025 policy change restated FY25 revenue to A$891.4 million and the FY26 accounts carry that basis. ROIC, net debt for FY22 and FY23, the owner cash bridge, all multiples, the scenario values and the FY27 NPAT model are author-computed from filed inputs and are not IDP-reported metrics. Volumes and average fees are IDP operational statistics. The market snapshot is the ASX close of 20 August 2026 cross-checked against a live market-data scan and a point-in-time price database; one commercial data vendor's market-capitalisation field disagreed with the filed share count, and the filed count of 278,336,211 shares was used. IDP is paid by universities for placements, which is a structural interest worth knowing when reading its commentary on student caps.