Bank of Montreal and REX Shares have added another high-leverage artificial-intelligence trading instrument to the U.S. exchange-traded market, launching the MicroSectors 3× Long MANGOS+ ex Private Companies ETNs under the ticker MNGU on NYSE Arca. The product began trading on August 27 following the companies’ August 26 launch announcement, giving sophisticated investors a single listed security designed to magnify the daily performance of a concentrated basket of companies positioned around the AI economy.

MNGU is linked to the net total return version of the NYSE MaNGoS+ Index, identified by the ticker MANGOSN. The notes seek three times the index’s performance from one daily leverage reset to the next, before the deduction of applicable fees and financing charges. That distinction is critical: MNGU is designed to target a 3X result for a single trading day, not three times the benchmark’s cumulative return over weeks, months or years.

The launch places MNGU at the intersection of two major exchange-traded product trends: persistent investor interest in artificial intelligence and growing product development around leveraged, tactical exposures. Instead of providing broad technology-sector exposure, the underlying benchmark concentrates on 10 companies selected under a rules-based methodology intended to capture both major AI platforms and companies tied to their commercial supply chains.

The index starts with six named “MaNGoS” anchor companies: Meta Platforms, Anthropic, Nvidia, Alphabet, OpenAI and Space Exploration Technologies, or SpaceX. Publicly listed anchors that meet the methodology’s eligibility rules can enter the index directly. Remaining positions are filled with qualifying public companies identified through disclosed customer or supplier relationships with the anchor group.

Anthropic and OpenAI are important exceptions. Both are included in the conceptual MaNGoS group but remained private as of the index’s launch and therefore are not constituents. MNGU consequently provides no initial economic exposure to either company. The methodology provides a route for a newly listed anchor to enter the index after satisfying eligibility requirements, potentially replacing a lower-ranked non-anchor constituent.

As of the start of August 27, the benchmark consisted of Palantir Technologies at approximately 13.39%, Microsoft at 12.91%, Nvidia at 10.09%, Meta at 10.00%, Taiwan Semiconductor Manufacturing at 9.52%, Alphabet at 9.26%, Advanced Micro Devices at 9.25%, Broadcom at 8.92%, Micron Technology at 8.86% and Space Exploration Technologies at 7.79%. Those weights move with market prices between scheduled index resets.

The index is reconstituted quarterly and assigns each constituent a 10% weight at reconstitution. That equal-weight framework distinguishes the benchmark from market-capitalization-weighted technology indexes in which the largest companies can dominate exposure. Even so, a 10-stock basket remains highly concentrated, particularly when its performance is magnified threefold on a daily basis.

Companies outside the anchor group face additional screening. According to the MicroSectors description of the methodology, the eligible universe includes common stocks and depositary receipts listed on major U.S. exchanges and is subject to thresholds including float-adjusted market capitalization, trading liquidity, free float and valuation requirements. Supply-chain relationships are used to define potential non-anchor candidates, after which an AI-relatedness calculation and ranking process further narrows the field.

The ranking methodology gives 35% weight each to full company market capitalization and six-month average daily traded value, while trailing price-to-sales and one-year net sales growth each account for 15%. The process therefore combines size and liquidity with valuation and growth variables rather than selecting companies solely because they are perceived by investors as AI-related.

For ETF and exchange-traded product investors, however, the most consequential feature is not the stock-selection process but the MNGU leverage mechanism. BMO’s offering documents emphasize that the notes reset their exposure each day. Because returns compound from one reset to the next, performance over multiple trading sessions can diverge sharply from three times the index’s cumulative move over the same interval.

That effect can work in either direction depending on the path of the benchmark. In a consistently rising market, daily compounding can sometimes produce a cumulative result greater than three times the index’s point-to-point gain. In a volatile market that repeatedly rises and falls, the same reset mechanism can erode value even if the benchmark ultimately finishes near its starting level. The impact generally becomes more important as volatility and holding periods increase.

BMO consequently describes the notes as daily trading tools for sophisticated investors rather than buy-and-hold investments. The issuer warns that MNGU can experience substantial decay and that investors holding the position for more than one day should continuously reassess the exposure. A sufficiently large adverse index move can inflict severe losses in a very short period because the underlying market movement is multiplied through the leverage structure.

The product is also structurally different from an exchange-traded fund. MNGU is an exchange-traded note, meaning it is a senior unsecured debt obligation of Bank of Montreal rather than a fund owning shares of the companies in the index. Investors therefore do not hold a proportional interest in an underlying securities portfolio and have no shareholder rights in the index constituents through the note.

Instead, payment obligations depend on BMO. That introduces issuer credit risk alongside the market, leverage and liquidity risks associated with the strategy. BMO’s pricing supplement states that payments on the ETNs are subject to the bank’s creditworthiness and that the notes are not deposits insured by U.S. or Canadian government deposit-insurance programs.

The initial principal amount is $25 per ETN. BMO’s pricing supplement covers an initial $25 million aggregate principal amount, representing 1 million notes. The issuer’s product page showed 1 million ETNs outstanding and a closing indicative note value of $26.64 as of August 27, corresponding to an indicative market capitalization of about $26.64 million.

The listed maturity date is July 31, 2046, although the long maturity should not be interpreted as an intended investment horizon. The offering documents repeatedly state that MNGU is not designed to be held to maturity. BMO also retains call rights under specified terms, while holders seeking issuer redemption face procedural conditions and minimum redemption requirements described in the prospectus.

Costs add another layer to the leverage mechanics. BMO and REX said the Daily Investor Fee is waived through January 31, 2027. After that period, the fee is scheduled to be based on an annual rate of 0.95%. The notes also incur a Daily Financing Charge, which the pricing supplement says is based on the Federal Reserve Bank Prime Loan Rate plus an initial spread of 3.00% per year. The spread can be increased under the product terms.

The pricing supplement also specifies a 0.125% redemption fee amount for holder-initiated early redemption unless waived by the issuer. Because financing expenses and investor fees are deducted from indicative value, the underlying index must generate enough favorable movement to overcome both those charges and any adverse compounding effect for investors to preserve or increase capital over a holding period.

That cost structure illustrates why headline leverage ratios do not fully describe the economics of leveraged ETNs. A trader may correctly predict the direction of an AI basket yet still realize a materially different result from a simple three-times calculation because of daily compounding, financing expenses, market-price premiums or discounts, liquidity conditions and the timing of entry and exit.

Liquidity is another consideration for a newly launched note. BMO states that an active secondary market is not guaranteed and that ETN trading prices can differ from indicative value. If issuance is limited or suspended while investor demand remains strong, the market price could trade above indicative value. A later collapse of such a premium could generate losses independent of the underlying index’s performance.

MNGU’s arrival also expands a MicroSectors lineup that has increasingly targeted narrow, high-conviction market themes. BMO and REX have launched leveraged products tied to semiconductors, artificial intelligence, corporate bonds and foreign equity markets, among other categories. The MANGOS+ product takes that specialization further by combining a branded group of AI leaders with supply-chain-linked companies selected under an index methodology.

For active traders, the appeal is straightforward. Rather than establishing and rebalancing leveraged positions across 10 individual equities, MNGU packages a rules-based basket into a single NYSE Arca-listed security. It also reduces the need to decide individually how much weight to allocate to chip designers, semiconductor manufacturers, cloud and software platforms and other companies exposed to AI capital spending.

The trade-off is concentration. The index’s constituents span different layers of the AI ecosystem, but their performance can still be driven by overlapping factors, including spending on data centers, semiconductor availability, enterprise AI adoption, capital expenditure plans, model economics and changes in valuations across growth stocks. A broad shift in expectations for AI investment could therefore affect several constituents simultaneously.

Three-times leverage magnifies that common-factor risk. A daily decline of several percentage points across large AI-related equities can translate into a substantially larger move in MNGU before expenses. The product disclosure goes further, warning that investors can lose their entire investment if the Intraday Indicative Value reaches zero or below at any time on an exchange business day, or if the Closing Indicative Value reaches zero.

The index itself also has only a short live operating history. ICE Data Indices launched the NYSE MaNGoS+ Index for live calculation on August 18, 2026. Historical figures before that date are back-tested and hypothetical, meaning they reflect the retroactive application of the methodology rather than the performance of a benchmark that investors could actually have tracked in real time.

For the exchange-traded product industry, MNGU underscores how issuers are moving beyond broad AI funds toward instruments designed for specific trading views. Traditional AI ETFs may hold dozens or hundreds of companies and are often positioned as portfolio allocations. MNGU instead concentrates exposure in 10 stocks and then applies daily 3X leverage, placing it firmly in the tactical end of the market.

The product also illustrates how index providers and issuers are trying to capture an evolving AI corporate ecosystem rather than define the theme through a static industry classification. By using disclosed supply-chain relationships and permitting future inclusion of newly public anchor companies, the MANGOS+ methodology can change as AI companies enter public markets or as relationships across the technology supply chain evolve.

That flexibility may become especially relevant if currently private AI developers eventually list their shares. Anthropic and OpenAI are part of the anchor framework but are absent from MNGU today. If either completes a qualifying public listing, the methodology provides for accelerated consideration rather than requiring investors to wait for a routine quarterly reconstitution. Such an event could materially alter the character of the benchmark.

For now, MNGU is primarily a leveraged expression of public-market AI infrastructure and platform exposure, with holdings spanning semiconductors, manufacturing, software, cloud platforms and other businesses tied to the development and deployment of artificial intelligence. Its launch gives traders another mechanism for making a concentrated bullish call on that ecosystem while adding the risks unique to daily leverage and unsecured bank-issued notes.

The key question for the product will be whether investor appetite for increasingly specialized AI exposures translates into sustained trading activity. Thematic investing has already moved from broad technology baskets toward narrower segments such as semiconductors, robotics, infrastructure and individual-stock leverage. MNGU pushes that progression into a curated AI network and multiplies the daily exposure threefold.

For investors evaluating the note, its branding is less important than its mechanics. MNGU is not a passive long-term AI fund, does not currently provide exposure to private companies such as OpenAI or Anthropic, and is not designed to deliver three times the benchmark’s long-term return. It is a short-horizon leveraged ETN whose results depend on daily index moves, compounding, financing costs, secondary-market conditions and the credit of Bank of Montreal. Those characteristics define both its potential usefulness as a trading instrument and the significant risks surrounding it.