Iran’s Venture Funds File the Same Note Eight Different Ways

Iran's venture funds file on Codal, and in most of their audited statements sits a note called summary financial information of venture investments. For the private companies inside those portfolios it is often the only place a revenue or profit figure appears in public at all, in any language. Tehran Index scored fifteen of those notes, from twelve funds, against five fields a reader needs before a number is usable. Two print all five, and both of those use the field to say the figures came from a tax return or an unaudited trial balance. Six of fifteen say where the numbers came from at all. Across the 49 investee rows that carry a period, the accounts run from the fund's own reporting date to twenty four months behind it.

Twenty venture funds file on Codal, the Iranian securities regulator’s disclosure system. In most of their audited statements there is a note with a heading that translates as summary financial information of venture investments: a table of the fund’s portfolio companies with each company’s own accounts printed beside its name. For the private companies inside those portfolios it is often the only place a revenue or a profit figure appears in public at all, in any language. Of the eleven Tehran Index records that carry one from this source, not one publishes accounts of its own.

Tehran Index opened 24 of those filings on 2 and 3 September 2026 and built two insights out of what the notes say. This one is about what they look like. We scored fifteen of them against five fields a reader needs before a figure is usable: revenue and net result, total equity, the period the figures cover, the basis they were prepared on, and a comparative.

Two of the fifteen print all five. Between them the fifteen produce eight different combinations of those five fields.

One thing has to be said before anything else. This is not a compliance finding and it is not a criticism of any filer. Tehran Index has not read the disclosure instruction the Securities and Exchange Organisation issues for venture funds, so we do not know what this note is required to contain, and we are not in a position to say that anyone has fallen short of it. Every filing scored here carries an auditor’s report. What we can say, because we read them, is that the same note in the same system produces materially different tables, and that the difference decides what an outside reader can do with the numbers.

The five fields

It is one disclosure. On the eight filings where the desk recorded the note’s Persian heading word for word, it is the same phrase every time. The note number is not: nine different numbers across the fifteen, the commonest being 10-2, which six filings use.

Revenue and a net result are nearly universal: fourteen of fifteen print both. The exception is Pishgam, which holds the largest portfolio in the whole sweep at eighteen positions and prints assets, liabilities and equity only. Eighteen companies, no income statement for any of them.

Total equity is also fourteen of fifteen, and the exception is instructive. The Pension Technology fund prints a column headed accumulated loss where every other filer prints total equity. Those are not the same line, and a reader who treats them as the same will foot a balance sheet that does not foot. The two rows in that table only reconcile once you infer the paid-in capital the filing never prints.

The field that separates a usable figure from an unusable one is the basis: whether the fund tells you that the investee numbers came from audited accounts, from a trial balance, or from a tax return. Six of fifteen do. Nine do not, and on those nine a reader has a revenue figure and no way to know what stands behind it.

The two filings that disclose most are the two admitting the weakest evidence

Both of the filings that complete all five fields do it by using the basis field to say the numbers are soft.

Arman Ati’s note 11-9-1 states that the latest-year figures for three of its investees, including Nobar and Pezeshk Khoob, were taken from the companies’ tax returns rather than from audited accounts. The University of Tehran’s second fund states that its single investee’s latest figures come from a trial balance and are unaudited. Neither disclosure flatters the filer. Both are the reason those two filings score full marks.

The two most complete disclosures in the set are complete because their authors wrote down the weakness. On the nine filings that leave the basis blank, the figures look stronger for no reason other than silence.

The same asymmetry runs through the whole table. Sarv, which carries nine companies, goes furthest of anyone: its annual prints an audit-status column per row and marks three of the nine as not audited or not stated. That column is the single most useful thing in any of these filings. Only Sarv prints it, and only in its annual; its own interim six months earlier does not.

The second thing that varies is how old the numbers are

A fund reports at its own year end. The investee accounts it prints are whatever the investee had ready. Nothing in any of these filings suggests the two dates are meant to meet, and they mostly do not.

Across the 49 investee rows that carry a printed period, the accounts run from the fund’s own reporting date to twenty four months behind it, with a median of nine months. Twenty of the 49 are a year old or more. Twelve carry the fund’s own date.

The spread inside a single table matters more than the average. Sarv’s annual prints nine companies at four different vintages, from three months behind its own year end to fifteen. Arghavan’s prints two companies side by side, one current and one fifteen months stale. Both funds print each company’s own year end in its own column, which is the only reason the spread is visible at all, and is the argument for that field rather than against those funds. Pishgam is the other extreme and deserves the credit: eleven of its fifteen dated rows are trial balances struck within two months of its own reporting date, which is why its note is the freshest in the sweep even though it prints no income statement at all.

Arman Ati’s table carries both ends. Three of its five rows are twelve months behind. The other two, Sahamyab and Azmayesh Online, are twenty four months behind, because the fund had nothing newer for them. It prints the old figures rather than a blank, which is the right call, and it dates them, which is the only reason anyone can tell.

What it does to our own record

This is not an abstract complaint about disclosure quality. It decides what this site can publish.

The Tehran Index registry holds 403 records, 376 of them public, carrying 1,181 public facts. Fifty six of those facts come from a listed venture fund’s filing, across fifteen records. Fifty two facts on the whole site carry a revenue or a net result, across thirty records, and twenty five of those fifty two come out of this one note.

They sit on eleven records: Basalam, Miare, Wallex, Dr. Saina, Khanoumi, Takhfifan, Iran Talent, Bimebazaar, Nobar, Pezeshk Khoob and Sahamyab. On every one of the eleven, every published revenue and profit figure comes from a shareholder’s note. Not one of the eleven publishes accounts of its own.

So for more than a third of the companies on this site that have a published income figure at all, the reliability of that figure is the reliability of this note, and the note is eight different things.

What it already cost two of our own pieces

Most of Iran’s Venture-Backed Companies Make a Profit, published on 9 September 2026, counted 28 companies across this sweep with a dated income statement and found 18 profitable in their latest reported period. That denominator is 28 and not larger for two reasons that are both in the scorecard above. Sepehr’s three investees were dropped because its note prints no period at all, so there is no latest period to count them in. Pishgam’s eighteen positions were never eligible at all, because it prints no income statement for any of them. Two funds’ choice of columns is most of the difference between that denominator and a larger one.

Six Iranian Venture Exits, Priced mined the same filings for what the funds sold and at what price. Both pieces read this note for its contents. This one is about the container, and the container is why the contents cannot simply be added up.

Two funds printed the same balance sheet and dated it a year apart

The clearest single illustration came out of the sweep by accident. Miare, a freight marketplace, has two listed venture funds on its register, and both filed interim statements to the same date, 21 December 2025. Both print Miare’s balance sheet. The three figures are identical to the rial.

Sarv dates them to the Iranian year end that falls on 20 March 2025. Pishgam dates the same three figures to the year end before it, 19 March 2024. The calendar settles it: the earlier of those two Iranian years had a twenty nine day final month, so the day Pishgam prints does not exist in the year it assigns. Sarv’s date holds.

That was catchable only because two funds happened to report the same company on the same day. Pishgam carries the same label on three other rows, including Basalam, and on those there is no second filer to check against. Our own Basalam record now carries the date the fund printed together with our reading of it and an open verification follow-up, rather than a clean date we cannot stand behind.

What we do not know

We have not read the disclosure instruction these funds file against. Without it, everything above is a measurement of variation and none of it is a judgement about compliance, and it should not be reported as one.

The set is what the desk could reach, not the population. Twenty funds file; seventeen had an audited letter on the first page of their filing history when the worklist was built on 2 September 2026. Three did not and are entirely unscored: Ashna Tek Iranian, Bamdad Atlas Yekom and Pouya Algorithm. Whether they carry audited statements deeper in their letter histories has not been established, and this piece does not assert that they do.

An interim and an annual are not the same document, and a no on an interim is not a statement about the fund. Firouzeh’s interim carries a narrative description of Ponisha and two other holdings instead of a table; its annual may well carry the table, and it has not been read. Stars publishes the note in the one annual of its three read letters that has it, and that copy is a scan too blurred to read digits from, so it is named and excluded rather than guessed at.

Six rows across the sweep do not foot: three in Sarv’s annual, one in Rooyesh Lotus’s IranTalent column, one Arman Ati comparative that prints an asset figure with a spare digit, and one in Pishraft’s interim, whose scan is coarse enough that the failure may be ours rather than the filing’s. Three further rows foot and are still held, because a balance sheet that adds up can still be irreconcilable with the column next to it: Rooyesh Lotus’s Bimeh Bazar comparative is about sixty four times its own later column, and Armani’s Vanda row foots while its equity roll-forward does not. None of the nine is published anywhere on this site. They are a separate problem from field presence and are counted in no rate above.

Every one of these 24 documents is a scan, or carries a text layer the desk could not use. What is scored is the desk’s reading of images, checked row by row against the accounting identity that assets equal liabilities plus equity, which is the strongest available proof that the digits were read correctly. Where a row failed that check it was held rather than published, and the holds are listed above.

And we do not know why any filer discloses less than another. That question is outside what the documents support and outside what this desk covers.

New in this piece, against the two that came before it: those counted what the notes say. This measures how the notes are built, across a fixed field set, and connects the shape of the disclosure to the exclusions the earlier counts had to make. If you take a figure out of an Iranian venture fund’s investee note, check four things before you use it: whether the period is printed, whether the basis is printed, how far behind the fund’s own date the accounts sit, and whether the row foots.

Cite as: Tehran Index, “Iran’s Venture Funds File the Same Note Eight Different Ways”, 20 September 2026. Scored from 15 of the 24 audited venture fund filings read on codal.ir on 2 and 3 September 2026, from a worklist of 20 funds built by filtering Codal’s full issuer list on the Persian for venture and for high risk. Registry figures computed from the Tehran Index company registry on 20 September 2026. tehranindex.com