Arbe Robotics: The Escrow Cash, Covenants, And The Nasdaq Linkage
This continuation isolates the financing layer the main article flagged: not all of Arbe's $24.5 million in escrow is accessible liquidity, and the real 2026 risk runs through the release mechanics, covenant wording, and the direct link between Nasdaq status and the bond deed.
What Is Actually Trapped
The main article argued that 2025 mostly bought Arbe time through financing. This continuation isolates the layer that is easiest to misread: not how much cash appears on the balance sheet, but how much of it is truly available to the company, under what conditions, and how that connects to both the covenants and Nasdaq.
The headline figure is $24.525 million of funds held in escrow at year-end 2025. But the number that has to sit next to it is $24.757 million of convertible bonds classified as a current liability. Those two lines nearly mirror one another. That is why the escrow balance is not a normal operating cash cushion. It is a protection layer sitting behind the bond series.
| Layer | Year-end 2025 balance | What it means in practice |
|---|---|---|
| Cash and cash equivalents | $4.028 million | Cash that is immediately available, but small on its own |
| Short-term bank deposits | $40.690 million | Available financial liquidity, and part of the covenant measurement after the amendment |
| Funds held in escrow | $24.525 million | Cash trapped with the trustee for the bondholders, not free operating liquidity |
| Convertible bonds, current liability | $24.757 million | The liability side of the same financing layer |
What matters is that the December 2025 private follow-on did not change the picture in the way many readers may assume. Arbe issued bonds with NIS 57.6 million of principal, but the proceeds that actually reached the escrow account were about NIS 50.2 million, or roughly $15.7 million. In other words, this structure expanded trapped cash well before it expanded operating freedom.
This chart matters because it separates two different stories. The first is that Arbe succeeded in adding more money to the bond framework. The second, and more important one, is that this money did not flow directly into the operating account. After the early 2025 conversions, $8.6 million remained in escrow, and the December follow-on lifted that amount to $24.3 million. In the year-end 20-F, the figure is already around $24.5 million. That is an addition to the collateral structure, not an open cash drawer.
The Release Map: Three Paths, Only One Really Creates Free Cash
The natural instinct is to talk about "releasing the escrow" as if there were one trigger. That is wrong. In practice there are three different paths, and they do not mean the same thing.
| Path | What has to happen | What Arbe gets in practice |
|---|---|---|
| Full release from escrow | By December 31, 2026, all three conditions have to be met together: a tender win or supply contract as the single supplier of imaging radar chips to one of ten named major automakers, an average Nasdaq closing price of at least $3.10 for 30 consecutive trading days, average combined Nasdaq and TASE volume of at least 300,000 shares per day, and a Nasdaq closing price of at least $3.10 on the day the company presents the documentation to the trustee | This is the path under which the cash can move to the company |
| Conversion of bonds into shares | Bondholders convert, and the related portion of escrow is released | Cash can move to the company, but only as long as the escrow balance left behind is not lower than the principal amount of the bonds still outstanding |
| Failure to meet the conditions by end-2026 | The company does not complete the release conditions on time | The escrow is not released to the company. It is used for early redemption of the bonds, and the company still has to fund interest and the U.S. dollar indexation differential as needed |
That is the key point. Conversion is not a magic source of fresh cash. It only releases the portion of escrow that is matched by a reduction in outstanding principal. So if a reader treats the escrow as another $24.5 million of available liquidity, they are missing the logic of the structure. This layer exists first to protect the bondholders, and only after that, under defined conditions, to create flexibility for Arbe.
The voluntary early redemption on January 16, 2026 makes that even clearer. The amendment to the deed gave holders a one-time option to choose early redemption after the extension and the interest-rate cut, and in practice bonds with NIS 836,842 of principal were redeemed for about $230 thousand. That was not another release to the company. It was a use of the escrow layer to reduce exposure for holders who opted out.
Covenants: The Real Story Is Disclosure Noise, Not Tight Headroom
The most confusing part of the disclosure is not the covenant level itself, but the way it is described. One description refers to a minimum cash and cash equivalents test of $5 million. If a reader stops there and compares it to the year-end balance sheet, they get $4.028 million, which seems to fall below the floor. That is too narrow a reading, because another document in the same legal package changes the definition.
The December 2025 amendment to the deed makes clear that the minimum cash balance test includes not only cash and cash equivalents, but also short-term bank deposits. Note 10 in the annual report says the same thing, describing the covenant as applying to cash and cash equivalents and short-term deposits, and it also states that the company was in compliance with all covenants as of the balance sheet date.
This chart captures the gap. If you insist on the narrow reading, Arbe looks close to the edge. If you read the amendment and the updated note, the picture flips: roughly $44.7 million against a $5 million liquidity floor, and about $39.6 million against the same $5 million equity floor. In other words, the year-end 2025 issue was not covenants that were getting tight. It was disclosure that was not speaking with one voice.
That matters because otherwise the reader can make two opposite mistakes. They can assume there is a covenant breach when there is no basis for that conclusion, or they can ignore the fact that the structure has become complicated enough to require a precise reading of the deed and its amendments, not just a management summary or a press release.
There is a second layer of noise here as well. The follow-on announcement refers to extending the "maturity date" to December 31, 2026, while the prospectus still states that the final maturity date of the bonds is May 30, 2028. What actually moved in December 2025 was not the final maturity. It was the deadline for satisfying the escrow release conditions, together with the interest-rate cut to 4.35% and the addition of the voluntary early redemption mechanism. That is not a minor wording issue. In a growth company buying time through financing, the timeline is part of the thesis.
Why Nasdaq Is Bond Risk, Not Just Equity Sentiment
The filings show a direct connection between Nasdaq status and the economics of the bonds across three separate layers.
The first layer is the release mechanism. Two of the three release conditions are built directly on Nasdaq: a $3.10 average closing price over 30 consecutive trading days, a minimum combined trading volume threshold, and a $3.10 closing price on the certification date. So even if the company progresses commercially, there is no full release of the escrow without a very strong equity-market window.
The second layer is listing compliance. In the annual report, Arbe states that as of the report date its Nasdaq share price had been below $1.00 per share since around February 25, 2026, creating a risk of a Nasdaq deficiency notice. That gap needs to be read correctly. The basic Nasdaq compliance floor is $1.00, but the escrow release threshold is $3.10. Put differently, even a return to ordinary listing compliance does not solve the escrow problem. It only avoids a worse one.
| Nasdaq layer | What is required or what can go wrong | Why it matters for the bonds |
|---|---|---|
| Basic listing compliance | Minimum bid price of $1.00 per share to remain listed | Delisting or suspension hurts market access and can worsen financing risk |
| Escrow release conditions | $3.10 average price for 30 trading days, combined volume of 300,000 shares per day, and $3.10 on the certification day | Without this, the cash does not open up to the company even if the series remains alive |
| Potential default event | The company states that Nasdaq delisting may constitute a default under the deed of trust | This is where Nasdaq stops being only an equity story and becomes part of the bond risk map |
The third layer is default risk. The annual report states explicitly that Nasdaq delisting or suspension of trading in the ordinary shares could also damage the market for the bonds and may trigger a default under the deed of trust. That is the point the market can miss if it looks at Nasdaq only as a matter of sentiment, liquidity, or valuation multiple. This is more structural than that: the bonds themselves trade only in Tel Aviv, yet the company documents tie the debt framework to the company's dual-listed status. Nasdaq is therefore part of the collateral and trigger architecture, not just part of the mood around the stock.
Bottom Line
The main article said that 2025 bought Arbe time. This continuation adds that the time came with very hard conditions. The $24.5 million in escrow is not an open cash balance. It is part of a mechanism that protects bondholders first. The covenants themselves do not look tight at year-end 2025, but uneven wording across the documents can hide that. At the same time, Nasdaq sits in the middle of the structure twice, once as a condition for fully releasing the cash and once as a trigger that can turn the bond from a bridge into a redemption story.
So the 2026 question here is not only commercialization. It is also whether the financing structure that stitched together 2025 can hold long enough for real operating economics to emerge. If not, the escrow will not become fuel for growth. It will simply go back to the bondholders.
Disclosure: Deep TASE analyses are general informational, research, and commentary content only. They do not constitute investment advice, investment marketing, a recommendation, or an offer to buy, sell, or hold any security, and are not tailored to any reader's personal circumstances.
The author, site owner, or related parties may hold, buy, sell, or otherwise trade securities or financial instruments related to the companies discussed, before or after publication, without prior notice and without any obligation to update the analysis. Publication of an analysis should not be read as a statement that any position does or does not exist.
The analysis may contain errors, omissions, or information that changes after publication. Readers should review official filings and primary sources before making decisions.