What Is Treasury Management and When Does a Startup Need It?
Treasury management governs a company's cash and financial risk. A funded startup needs it from its first round, as a board-approved investment policy.
Treasury management is the function that keeps a company able to pay, invests the cash it does not yet need and manages the financial risks that come with both. A funded startup needs it from its first priced round, first as a board-approved investment policy. A treasurer, hedging and software come later, with currencies, debt or customer funds.
The order follows from what a startup’s balance sheet is. Tim Opler, Lee Pinkowitz, René Stulz and Rohan Williamson studied US public companies from 1971 to 1994. Firms, they found, hold liquid assets so they can keep investing when cash flow falls short and outside money is expensive, and large swings in excess cash come mainly from “the occurrence of operating losses” (Opler et al., 1999). A venture-backed company, which funds its losses from cash raised in rounds, fits that description more starkly than the public companies studied. Venture lenders can also make the investment policy a contract term: the loan agreements that Couchbase (2021) and HawkEye 360 filed with the SEC permit investments beyond cash equivalents only if the policy allows them and the lender has approved it in writing.
The jobs of treasury management, and when a startup takes each on
| Job | The decision | Startup trigger | Where it is written down |
|---|---|---|---|
| Cash visibility | What the company holds, where, and what it can spend today | First bank account | A daily cash report |
| Liquidity | How many months of burn stay available the same day | First priced round | The investment policy’s operating tier |
| Investing the reserve | Permitted instruments, maturity limits, issuer limits, credit floor | First priced round | The investment policy, approved by the board |
| Counterparties | Which banks, funds, brokers and custodians may hold the cash | A second bank or a first brokerage account | The policy’s approved list |
| Debt | What to borrow, against which covenants, and which investments the lender permits | First loan | The loan agreement |
| Market risk | Whether to hedge currencies and interest rates, and how to account for hedges | Material costs or revenue in a second currency; floating-rate debt | A hedging policy and ASC 815 documentation |
Source: AFP glossary for the jobs; loan agreements and filings cited below. The triggers are CX Cash’s judgement.
What does treasury management cover?
Cash, liquidity and financial risk, decided before the money moves. The Association for Financial Professionals, which certifies corporate treasurers, defines treasury management as “the process of overseeing a company’s financial resources (including cash, assets and liabilities) to achieve the company’s strategic objectives”. Its core activities run from liquidity and investing excess funds, “prioritizing principal preservation, liquidity and overall return”, to risk and financing (AFP glossary). Cash management is the daily subset: balances, payments and the short forecast.
The word has a second meaning. Banks sell “treasury management” as a product line: Falcon National Bank describes it as “a suite of specialized banking services” covering payments, collections, fraud controls such as Positive Pay and account reporting (Falcon National Bank, 2025). AFP’s glossary also records a choice that matters later: treasury can run as a cost centre or as a profit centre expected to earn from trading, hedging or investment. Of the second model it warns: “The potential downside is the risk associated with the pressure to produce profits from these activities.”
Why does a funded startup need treasury management so early?
Because after a round almost all of its balance sheet is precautionary cash, and studies of public companies show such cash growing with risk. Thomas Bates, Kathleen Kahle and René Stulz found that the average cash-to-assets ratio of US industrial firms more than doubled from 1980 to 2006. “Cash ratios increase because firms’ cash flows become riskier”, they wrote, and the precautionary motive played an important role (Bates, Kahle & Stulz, 2009). In their working paper, the median cash ratio of firms within five years of an IPO tripled over the sample period (NBER w12534, 2006). A software company can stay at that end even after it matures: Samsara reported revenue of $1,618.6 million for the year to 31 January 2026, yet cash, cash equivalents and investments made up 48.7% of its total assets (Samsara 10-K).
A Series A software company typically has no inventory, few receivables and little operating cash flow, so its treasury is a stock of cash that operating losses shrink every month. If the product moves customers’ money, their funds sit in the same accounts, so burn shows only once each flow is tagged as the company’s or a client’s. In one card-payments scale-up’s experience, a treasury failure meant cards declined at the till, which justified an experienced treasurer before finance had accounting software. Either way, a loss on the reserve comes straight out of runway, and a company without operating profit cannot earn it back, which is why the decisions about it belong in a written investment policy.
What goes into a startup’s investment policy?
Five answers: permitted instruments, a maturity limit, a concentration limit, a credit floor and who approves changes. Public companies describe their policies in a sentence or two, while universities and local governments publish theirs in full, which makes them a startup’s best templates.
What five investment policies say
| Policy | Permitted instruments | Maturity limit | Concentration limit | Credit floor | Who approves or reviews |
|---|---|---|---|---|---|
| SentinelOne, 10-K for the year to 31 January 2026 | Highly rated securities | Weighted average of 18 months or less | Limits exposure to any one issuer | Investment grade | Not disclosed |
| AppDynamics, S-1, 2016 | Investment-grade marketable securities | Two years maximum; average under one year | No significant concentration in any one issuer | Investment grade | Not disclosed |
| DoubleVerify, 10-K for 2025 | Treasury bills and notes, money market funds, savings accounts | Implied by the instruments | Not stated | Not stated | Not disclosed |
| University of California Short Term Investment Pool, effective 1 July 2025 | Treasury and agency notes, CDs, commercial paper and others | Three years remaining maturity | 3% per issuer outside Treasuries and agencies; 5% of the commercial paper allocation per issuer | A or better; commercial paper A-1, P-1 or F-1 | Board of Regents approves changes |
| University of Arkansas operating funds | Treasuries, agencies, municipals, corporates, commercial paper, CDs, repos, money funds | Tier 1: 12 months; Tier 2: 5 years; Tier 3: 10 years; commercial paper 270 days | Corporate notes 35% in total, 5% per issuer | Commercial paper A-1/P-1; corporates A-/A3 | Committee reviews at least annually |
Source: SentinelOne 10-K; AppDynamics S-1; DoubleVerify 10-K; UC Regents Policy 6109; University of Arkansas Policy 3140.
The university documents add the order of priorities. Arkansas ranks its objectives as safety of principal, then liquidity, then return, and splits its funds into tiers by when the money will be needed. Part of Tier 1 stays in money market funds, deposits or overnight repos because of “the difficulties inherent in accurately forecasting cash flow requirements”. A startup’s equivalent tiers are measured in months of burn.
California writes the governance into law. Under its Government Code, a county treasurer may render an annual statement of investment policy, “which the board shall review and approve at a public meeting”. A quarterly report under the same section must state whether the portfolio complies and whether the agency can meet its pool’s “expenditure requirements for the next six months” (Cal. Gov. Code § 53646). The same six-month test suits a startup’s board report, since it ties the portfolio to the spending it must cover.
What goes wrong when treasury has no written limits?
Treasury fails from both ends: the pressure to earn, and a wrong judgement that something is cash. Robert Citron, Orange County’s elected Treasurer-Tax Collector, won re-election seven times (Wikipedia). He had borrowed against the county’s $8 billion of pooled funds in the repo market, leveraging them to between 158% and more than 292%. When the Federal Reserve raised rates in 1994, dealers demanded margin, and the county filed for Chapter 9 bankruptcy on 6 December 1994, with $1.64 billion of losses and about 3,000 jobs cut (Wikipedia). Citron, it later emerged, had taken his interest-rate forecasts from a mail-order astrologer and a psychic. The county’s pool was AFP’s profit-centre model at its extreme.
Professional judgement fails too, so a policy should define cash by how it comes back rather than by its name. Auction-rate securities were long-dated bonds whose rates reset at periodic auctions, and by 2007 they formed a $330 billion market. In June 2005 the head of AFP said that “auction rate securities have proven to be highly liquid investments” (Congressional Research Service, 2012). On 13 February 2008 most major dealers stopped supporting the auctions, and more than 85% of the market saw failures in mid-February. One survey of filings found 430 non-financial companies holding $21 billion of them at par; 395 had marked them down, typically by 15% to 30%. At a startup, frozen paper could be most of the cash. One startup accountant recalls a biotech client that kept the bulk of its reserve in paper sold to founders as higher-yielding cash, and collapsed.
Which limits protect the most runway?
The concentration limit and the definition of cash, once each limit is restated as the months of runway the company would lose if the event it guards against happened. The table below does that for a labelled hypothetical.
Suppose a company holds $18.0M after its Series A and burns $600k a month, 30 months of runway. Its policy keeps 3 months of burn ($1.8M) in deposits at two banks, and the next 9 months ($5.4M) in a government money fund and Treasury bills of up to 12 months. The remaining $10.8M is a reserve of Treasuries, agency notes, A-1/P-1 commercial paper and corporate notes rated A or better. Its limits are a 24-month maximum maturity, a weighted-average maturity of 12 months or less and 5% of the portfolio per non-government issuer.
Each policy limit restated in months of burn, for a hypothetical company after its Series A, $
| Event | Limit that bounds the loss | Loss | Months of burn |
|---|---|---|---|
| Rates rise 1 point | Weighted-average maturity: 12 months (reserve), about 6 months (liquidity tier) | 135,000 | 0.23 |
| Rates rise 1 point, reserve in 5-year notes | No maturity limit (upper bound) | 567,000 | 0.94 |
| One issuer defaults, nothing recovered | 5% per issuer | 900,000 | 1.50 |
| The same, at the UC pool’s limit | 3% per issuer | 540,000 | 0.90 |
| The same, with the whole corporate allowance in one name | No issuer limit; 35% sector cap | 6,300,000 | 10.50 |
| A “cash equivalent” holding 20% of the portfolio freezes | The list of permitted instruments | 540,000 to 1,080,000 marked down; 3,600,000 locked up | 0.9 to 1.8; 6.0 locked up |
| The same, holding everything outside the deposits (90%) | The list of permitted instruments | 2,430,000 to 4,860,000 marked down; 16,200,000 locked up | 4.1 to 8.1; 27.0 locked up |
Source: CX Cash calculation on a hypothetical company. Rate losses use the weighted-average maturity as an upper bound for duration (loss ≈ amount × years × 1%). Markdown range: CRS (2012), typical auction-rate markdowns of 15% to 30%. Issuer limits: the five policies above.
The approximation is conservative. SentinelOne caps its weighted-average maturity at 18 months, and it reported that a 100 basis point change in rates would move its portfolio’s fair value by $3.3 million: 0.55% of $599.9 million of investments, an implied duration of about half a year. At the hypothetical company’s limits, a full point on rates costs $135,000, about 6.8 days of burn (our arithmetic). A single issuer’s default at the 5% limit costs 6.7 times as much, and an unlimited concentration costs nearly a year. The corporate policies above spend most of their words on maturities and ratings. One cash manager that serves startups reverses both of the hypothetical’s choices, capping maturities at just over a year but buying corporates down to BBB+. At these limits, that saves days and risks months.
When does a startup need hedging?
When it has a material cost or revenue line in a second currency, or debt at a floating rate. That is where the risk half of treasury management begins, and size alone does not decide it. US accounting made hedge accounting optional from the start: FASB’s Statement 133, issued in 1998, provided “special optional hedge accounting”, and a 2017 overhaul of its successor, ASC 815, applies to any entity “that elects to apply hedge accounting” (FASB, ASU 2017-12).
Three stages of the risk half of treasury, from filings
| Stage | Company and filing | What it says |
|---|---|---|
| No hedging | Samsara, 10-K for the year to January 2026 ($1,618.6M revenue) | Has not entered into any hedging arrangements for currency risk |
| No hedging | Procore, 10-K for 2025 | Has not hedged because currency effects were not material, but may if exposure “becomes more significant” |
| Weighing a programme | Instructure, S-1, 2015 | Evaluating the costs and benefits of a hedging programme as international operations grow |
| Economic hedges | SentinelOne, 10-K for the year to January 2026 | Forwards on an Israeli shekel liability, $154.4M notional, not designated as accounting hedges |
| Designated hedges | Zscaler, 10-K for the year to July 2026 | Forwards up to 24 months designated as cash flow hedges; swaps on $750.0M of its 2028 convertible notes designated as fair value hedges |
Source: Samsara, Procore, Instructure, SentinelOne and Zscaler filings.
Samsara had $1.6 billion of revenue, with a substantial majority of its sales in dollars, and no currency hedges; SentinelOne, with mostly dollar contracts, hedges a liability in Israeli shekels. Fastly’s pre-IPO loans carried interest at prime plus 1.75% and prime plus 4.25% (Fastly S-1, 2019), so every move in the prime rate passed straight into its interest bill. Two rules soften the start: smaller reporting companies are exempt from the SEC’s market-risk disclosure (17 CFR 229.305), and ASU 2017-12 gives private companies that are not financial institutions extra time for their hedge-effectiveness assessments.
Is a written policy worth it if the company only holds Treasury bills?
Yes, because its value lies in what it forbids and in who may change it. The case against is that most startups’ answer is short and obvious. DoubleVerify’s policy keeps its surplus funds in “treasury bills, treasury notes, money market funds and savings accounts to reduce its exposure to market risk”. For a company of 20 people, a written document and a board vote can look like ceremony.
On the instruments, that case holds. For most seed and Series A companies the list should be short, and a policy naming Treasury bills, government money funds and deposits at two banks is a good one; nothing in the months-of-burn table rewards reaching for yield. An extra half point on the hypothetical’s cash earns about $112,500 over its 30-month run-down, 5.6 days of burn. One issuer’s default at the 5% limit costs 8 times as much (our arithmetic).
But the pull toward yield grows with the round: the same cash manager finds that companies with the longest runways manage the cash they never expect to touch for return. Many buyers of auction-rate securities thought they held cash equivalents, and the same table shows that the costly failures are concentration and misclassified cash, which a short list does not by itself prevent.
The lender will ask for the document anyway. Bridge Bank’s 2009 loan agreement with Pandora allowed other investments only as “permitted under Borrower’s investment policy as approved by Borrower’s board of directors” (Pandora S-1/A exhibit). Lenders usually hold the cash too: HawkEye 360’s mezzanine loan requires at least 75% of it, and Couchbase’s 2021 agreement kept its primary accounts at Silicon Valley Bank. When that bank failed in March 2023, some finance chiefs said they had to choose between leaving cash at a failing lender and breaching their loan to move it. A second bank must be written into the loan before signing.
Treasury management at a startup therefore begins at the first priced round, as decisions about a shrinking pile of precautionary cash, written into a policy the board approves and the lender can read. A treasurer, hedging and software can wait until a second currency, debt or customer funds arrive. The first decisions carry uneven stakes: at the hypothetical company’s limits, a full point on rates costs about a week of burn, and an unlimited concentration in one issuer costs nearly a year.
Sources and notes
- Association for Financial Professionals. Treasury Management (glossary).
- Bates, T. W., Kahle, K. M. and Stulz, R. M. (2009). Why do U.S. firms hold so much more cash than they used to? Journal of Finance 64(5), 1985–2021; NBER Working Paper 12534 (2006).
- California Government Code § 53646.
- Congressional Research Service (2012). Auction-Rate Securities, RL34672, updated 17 July 2012.
- Falcon National Bank (2025). What Is Treasury Management and Why Does Your Business Need It?
- Financial Accounting Standards Board (2017). Accounting Standards Update 2017-12, Derivatives and Hedging (Topic 815): Targeted Improvements to Accounting for Hedging Activities.
- Opler, T., Pinkowitz, L., Stulz, R. and Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of Financial Economics 52(1), 3–46; NBER Working Paper 6234 (1997).
- Regulation S-K, Item 305, 17 CFR 229.305.
- SEC filings: AppDynamics S-1 (2016); Couchbase S-1, Exhibit 10.13 (2021); DoubleVerify 10-K (2025); Fastly S-1 (2019); HawkEye 360 S-1, Exhibit 10.12 (2026); Instructure S-1 (2015); Pandora Media S-1/A, Exhibit 10.16 (2011); Procore 10-K (2025); Samsara 10-K (fiscal 2026); SentinelOne 10-K (fiscal 2026); Zscaler 10-K (fiscal 2026).
- University of Arkansas. Fayetteville Policy 3140, Investment Policy – Operating Funds.
- University of California Board of Regents. Policy 6109, Short Term Investment Pool Investment Policy Statement, effective 1 July 2025.
- Wikipedia. Robert Citron; Orange County, California.
Company facts come from the SEC filings listed, the policy terms from the university policies and California’s code, and the Orange County and auction-rate histories from Wikipedia and the Congressional Research Service. Where a figure has no source of its own, or is marked “our arithmetic”, it is our calculation on the cited filings or on the stated assumptions of the months-of-burn table, whose company is hypothetical.
More in Cash & Treasury Management
Treasury Management Software: Where a Startup's Cash Is Held
Treasury management software for startups, compared: Mercury, Brex, Rho, Ramp and five TMS vendors, judged as treasurers rank them: safety, liquidity, yield.
Why Cash Flow Visibility Software Runs Hours Behind the Bank
Cash flow visibility software can show nothing newer than the bank's last report. How balances age, which status to trust, and what a stale number costs.
How Much to Hold and Where to Park Idle Startup Cash
Where to park idle startup cash: hold runway to the next milestone plus a raise, fence the surplus, then choose each home by its exit on a bad day.