MicroSectors FANG Innovation 3x Leveraged ETN (NYSEARCA: BULZ) closed at $41.05 on Friday, June 5, 2026, down roughly 18% from the prior session’s close of $49.89.
A holder with $10,000 in BULZ at Thursday’s close woke up Saturday with approximately $8,228, the direct result of a product designed to triple daily index moves doing precisely that.
The week-long picture is even more stark, with BULZ falling from $53.98 on May 29 to $41.05 on June 5, a drawdown of roughly 24% across just five trading sessions.
Despite the carnage, the note remains up approximately 192% over the trailing twelve months and around 58% year to date, which explains why it still commands a significant retail following.
BULZ is an exchange-traded note issued by Bank of Montreal, meaning holders carry BMO’s senior unsecured credit risk in addition to any movement in the underlying index.
The note tracks three times the daily return of the NYSE FANG+ Innovation Index, a ten-name basket that includes NVIDIA (NASDAQ: NVDA), Tesla (NASDAQ: TSLA), Amazon (NASDAQ: AMZN), Meta Platforms (NASDAQ: META), and Broadcom, alongside Alphabet, Netflix, and comparable names.
The 3x daily reset re-strikes exposure every morning to deliver three times that single day’s index return, not three times the return since purchase, which is the mechanism that both built the twelve-month gain and produced Friday’s loss.
That same structure also drives volatility decay, a property that has left BULZ up only 0.22% over the trailing month even as headlines around its holdings remained broadly constructive.
Friday’s selloff was coordinated across the basket, with NVIDIA falling roughly 6% to $205, Broadcom dropping approximately 8% to $386, Tesla retreating about 7% to $391, Meta sliding roughly 6% to $593, and Amazon declining around 3% to $246.
Run those moves through an equal-weighted basket and the index posted a high-single-digit loss for the day, which multiplied by three produces an 18% outcome with no tracking error or issuer fault involved.
Two catalysts stacked to break the tape this week rather than last, the first being Broadcom’s earnings report, where CEO Hock Tan delivered $22.19 billion in revenue and $2.44 in non-GAAP EPS against a $2.40 consensus, with AI semiconductor revenue hitting $10.8 billion, up 143% year over year.
The Q3 guide of $29.4 billion in total revenue and $16.0 billion in AI semiconductor revenue beat in absolute terms but fell short of the bar traders had already priced in, sending Broadcom down roughly 13% to 15% and spreading the pressure across the AI complex.
The second catalyst was Friday’s May payrolls print, which came in at 172,000 against an 80,000 estimate, yanking the two-year Treasury yield to 4.16%, a sixteen-month high, and reviving rate-hike conversation the market had largely dismissed.
Alphabet’s announced $80 billion stock sale earlier in the week to fund AI capital expenditure also sharpened the question of how much these companies must keep spending to justify current valuations, giving the selloff a coherent narrative.
What stands out is how contained the broader market reaction remained, with the VIX closing Thursday at 15.40, sitting in the 15.6th percentile of its trailing twelve-month range, well below the 31.05 peak recorded on March 27, 2026.
An 18% down day in a 3x leveraged product requires roughly a 22% up day to recover, meaning the index would need to deliver something close to 7% in a single session to get BULZ back to Thursday’s close.
The next key data point is the CPI print on June 11, where a hot reading stacked on top of the May payrolls surprise could push the two-year yield through 4.20% and re-rate the FANG+ basket again.
Fed speakers responding to the May jobs report will signal how seriously the committee is treating the apparent reacceleration, while any updated capital expenditure framing from the hyperscalers will determine whether AI spending continues to be celebrated or increasingly punished.
The mechanism that made BULZ a 192% twelve-month winner and the mechanism that made it an 18% one-day loser are identical, and neither outcome represents the product malfunctioning.
If the AI capital expenditure cycle keeps expanding and rates drift lower the leverage compounds favorably, but if May’s jobs number marks the start of a genuine reacceleration the path-dependency works the other direction, and year-to-date gains can disappear faster than they accumulated.
