STOCK//SHORTERSUBSCRIBE — FREE

STOCK//SHORTER / BACKTEST ARCHIVE

Historical AI disruption backtests.

Reconstructed as-of theses for AI-exposed software names, ranked by a forward downside model that weighs AI substitution, revenue compression, debt/cash-flow fragility, evidence depth, failure horizon, and strategy fit. The archive is intentionally separate from the blog and newsletter.

Disclosure: These are historical reconstructions, not posts that were published on the as-of dates. Research only, not investment advice, and not a recommendation to buy or sell any security.
Records22
Average Conviction69/100
Debt/Cash High Fragility6
PublishedJune 19, 2026
Forward 12-Month Downside Model

Top 10 AI disruption short candidates.

#1 · FORWARD DOWNSIDE 84/100 · CHGG

Short CHGG (Chegg)

Reconstructed as-of thesis: study-help subscriptions faced direct AI substitution as students moved from paid answer libraries to general-purpose tutors.

AI Conviction 91Debt/Cash 82Failure 6-18 months if churn and cash generation keep deteriorating.

#2 · FORWARD DOWNSIDE 81/100 · RNG

Short RNG (RingCentral)

Reconstructed as-of thesis: communications seats and contact-center workflows face AI-agent substitution while leverage can amplify downside in a demand shock.

AI Conviction 86Debt/Cash 83Failure 12-30 months if leverage, refinancing cost, and revenue compression collide.

#3 · FORWARD DOWNSIDE 76/100 · ASAN

Short ASAN (Asana)

Reconstructed as-of thesis: AI work orchestration can replace parts of manual project management and compress collaboration-tool seats.

AI Conviction 81Debt/Cash 74Failure 12-24 months if cash burn persists and revenue growth decelerates.

#4 · FORWARD DOWNSIDE 75/100 · FVRR

Short FVRR (Fiverr)

Reconstructed as-of thesis: AI commoditizes low-end creative, copy, coding, and support tasks that drove marketplace volume.

AI Conviction 84Debt/Cash 67Failure 12-24 months for model impairment if spend fails to migrate to higher-value services.

#5 · FORWARD DOWNSIDE 73/100 · BILL

Short BILL (BILL Holdings)

Reconstructed as-of thesis: SMB finance automation faces AI-native bookkeeping, payments, and AP agents that can compress standalone workflow software spend.

AI Conviction 78Debt/Cash 68Failure 12-30 months if AI substitution meets SMB revenue softness and credit-cycle pressure.

#6 · FORWARD DOWNSIDE 72/100 · UPWK

Short UPWK (Upwork)

Reconstructed as-of thesis: AI reduces transaction volume for repeatable freelance labor while enterprise buyers scrutinize contingent workforce budgets.

AI Conviction 79Debt/Cash 61Failure 12-30 months, depending on enterprise resilience and AI-service marketplace offset.

#7 · FORWARD DOWNSIDE 70/100 · PATH

Short PATH (UiPath)

Reconstructed as-of thesis: agentic AI threatens rules-based robotic process automation by replacing brittle scripts with flexible task agents.

AI Conviction 82Debt/Cash 48Failure 12-24 months for demand impairment if agent platforms take budget share.

#8 · FORWARD DOWNSIDE 70/100 · NCNO

Short NCNO (nCino)

Reconstructed as-of thesis: bank workflow software faces AI-native loan origination, credit, and relationship-management automation while financial customers scrutinize software budgets.

AI Conviction 74Debt/Cash 66Failure 12-30 months if bank tech budgets tighten and AI-native automation reduces workflow-seat demand.

#9 · FORWARD DOWNSIDE 68/100 · CXM

Short CXM (Sprinklr)

Reconstructed as-of thesis: AI-native customer-service and social-listening tools pressure large CX software suites.

AI Conviction 76Debt/Cash 56Failure 12-24 months if growth deceleration and margin pressure continue together.

#10 · FORWARD DOWNSIDE 63/100 · DBX

Short DBX (Dropbox)

Reconstructed as-of thesis: file storage and collaboration features face AI search, bundled productivity storage, and workflow abstraction.

AI Conviction 69Debt/Cash 63Failure 18-36 months if mature cash flow cannot offset revenue stagnation and leverage.
Strategy Backtest Queue

Highest priority strategy scenarios.

#1CHGGLong put research scenario83/100#2RNGLong put research scenario82/100#3ASANLong put research scenario79/100#4PATHLong put research scenario76/100#5BILLLong put spread research scenario75/100#6UPWKLong put spread research scenario75/100#7NCNOLong put spread research scenario74/100#8CHGGShort equity research scenario70/100

Strategy scores rank research/backtest priority only. Verified returns require historical prices, borrow, dividends, spreads, option chains, and execution assumptions.

Short equityentry price, exit price, borrow cost, dividend costuncapped loss risk
Long putstrike, entry premium, exit underlying or exit premium, expirationpremium paid
Long put spreadlong/short strikes, premiums, exit underlying, expirationnet debit
Bear call spreadshort/long strikes, premiums, exit underlying, expirationspread width minus net credit
Short callstrike, premium, exit underlying, explicit margin/capital-at-riskuncapped loss risk
Full Coverage Set

20+ reconstructed thesis records.

#1 · AS-OF June 21, 2024 · CHGG

Short CHGG (Chegg)

Reconstructed as-of thesis: study-help subscriptions faced direct AI substitution as students moved from paid answer libraries to general-purpose tutors.

Forward 84Conviction 91Debt/Cash 82Evidence 70
READ THE RECONSTRUCTED THESIS ->

#2 · AS-OF January 16, 2026 · RNG

Short RNG (RingCentral)

Reconstructed as-of thesis: communications seats and contact-center workflows face AI-agent substitution while leverage can amplify downside in a demand shock.

Forward 81Conviction 86Debt/Cash 83Evidence 59
READ THE RECONSTRUCTED THESIS ->

#3 · AS-OF June 20, 2025 · ASAN

Short ASAN (Asana)

Reconstructed as-of thesis: AI work orchestration can replace parts of manual project management and compress collaboration-tool seats.

Forward 76Conviction 81Debt/Cash 74Evidence 57
READ THE RECONSTRUCTED THESIS ->

#4 · AS-OF July 19, 2024 · FVRR

Short FVRR (Fiverr)

Reconstructed as-of thesis: AI commoditizes low-end creative, copy, coding, and support tasks that drove marketplace volume.

Forward 75Conviction 84Debt/Cash 67Evidence 62
READ THE RECONSTRUCTED THESIS ->

#5 · AS-OF November 21, 2025 · BILL

Short BILL (BILL Holdings)

Reconstructed as-of thesis: SMB finance automation faces AI-native bookkeeping, payments, and AP agents that can compress standalone workflow software spend.

Forward 73Conviction 78Debt/Cash 68Evidence 55
READ THE RECONSTRUCTED THESIS ->

#6 · AS-OF August 16, 2024 · UPWK

Short UPWK (Upwork)

Reconstructed as-of thesis: AI reduces transaction volume for repeatable freelance labor while enterprise buyers scrutinize contingent workforce budgets.

Forward 72Conviction 79Debt/Cash 61Evidence 60
READ THE RECONSTRUCTED THESIS ->

#7 · AS-OF March 21, 2025 · PATH

Short PATH (UiPath)

Reconstructed as-of thesis: agentic AI threatens rules-based robotic process automation by replacing brittle scripts with flexible task agents.

Forward 70Conviction 82Debt/Cash 48Evidence 61
READ THE RECONSTRUCTED THESIS ->

#8 · AS-OF March 20, 2026 · NCNO

Short NCNO (nCino)

Reconstructed as-of thesis: bank workflow software faces AI-native loan origination, credit, and relationship-management automation while financial customers scrutinize software budgets.

Forward 70Conviction 74Debt/Cash 66Evidence 49
READ THE RECONSTRUCTED THESIS ->

#9 · AS-OF April 18, 2025 · CXM

Short CXM (Sprinklr)

Reconstructed as-of thesis: AI-native customer-service and social-listening tools pressure large CX software suites.

Forward 68Conviction 76Debt/Cash 56Evidence 54
READ THE RECONSTRUCTED THESIS ->

#10 · AS-OF September 19, 2025 · DBX

Short DBX (Dropbox)

Reconstructed as-of thesis: file storage and collaboration features face AI search, bundled productivity storage, and workflow abstraction.

Forward 63Conviction 69Debt/Cash 63Evidence 50
READ THE RECONSTRUCTED THESIS ->

#11 · AS-OF December 19, 2025 · TWLO

Short TWLO (Twilio)

Reconstructed as-of thesis: AI customer-engagement agents can reduce usage-based communications growth and pressure standalone messaging infrastructure economics.

Forward 59Conviction 70Debt/Cash 45Evidence 51
READ THE RECONSTRUCTED THESIS ->

#12 · AS-OF January 17, 2025 · CRM

Short CRM (Salesforce)

Reconstructed as-of thesis: CRM was a high-multiple seat-based software incumbent facing AI-agent pressure on workflow automation, sales operations headcount, and renewal seat expansion.

Forward 58Conviction 72Debt/Cash 34Evidence 58
READ THE RECONSTRUCTED THESIS ->

#13 · AS-OF February 20, 2026 · WK

Short WK (Workiva)

Reconstructed as-of thesis: AI reporting, audit, and compliance agents can pressure premium workflow seats for collaborative financial and regulatory reporting.

Forward 57Conviction 62Debt/Cash 52Evidence 46
READ THE RECONSTRUCTED THESIS ->

#14 · AS-OF February 21, 2025 · DOCU

Short DOCU (DocuSign)

Reconstructed as-of thesis: e-signature becomes a bundled workflow feature while AI contract agents attack document preparation and agreement management.

Forward 56Conviction 68Debt/Cash 38Evidence 52
READ THE RECONSTRUCTED THESIS ->

#15 · AS-OF December 20, 2024 · ZM

Short ZM (Zoom)

Reconstructed as-of thesis: AI meeting summaries and bundled collaboration suites threaten standalone communications pricing.

Forward 51Conviction 66Debt/Cash 22Evidence 53
READ THE RECONSTRUCTED THESIS ->

#16 · AS-OF November 15, 2024 · ADBE

Short ADBE (Adobe)

Reconstructed as-of thesis: generative creative tools pressure Adobe's pricing power, but the incumbent has distribution, workflow lock-in, and its own AI products.

Forward 49Conviction 63Debt/Cash 28Evidence 56
READ THE RECONSTRUCTED THESIS ->

#17 · AS-OF August 15, 2025 · BOX

Short BOX (Box)

Reconstructed as-of thesis: enterprise content management faces AI-search and workflow abstraction, though mature cash generation can reduce downside urgency.

Forward 49Conviction 59Debt/Cash 43Evidence 45
READ THE RECONSTRUCTED THESIS ->

#18 · AS-OF September 20, 2024 · DOCS

Short DOCS (Doximity)

Reconstructed as-of thesis: pharma and healthcare marketing budgets may shift toward AI-driven targeting and lower-cost digital channels.

Forward 47Conviction 57Debt/Cash 24Evidence 47
READ THE RECONSTRUCTED THESIS ->

#19 · AS-OF May 16, 2025 · WDAY

Short WDAY (Workday)

Reconstructed as-of thesis: AI agents threaten HR and finance workflow seats, but enterprise system-of-record lock-in reduces failure probability.

Forward 43Conviction 55Debt/Cash 26Evidence 46
READ THE RECONSTRUCTED THESIS ->

#20 · AS-OF October 17, 2025 · PAYC

Short PAYC (Paycom)

Reconstructed as-of thesis: payroll and HR workflow automation faces AI self-service and suite competition, but mission-critical retention reduces failure probability.

Forward 42Conviction 54Debt/Cash 25Evidence 43
READ THE RECONSTRUCTED THESIS ->

#21 · AS-OF July 18, 2025 · MNDY

Short MNDY (monday.com)

Reconstructed as-of thesis: AI work agents can reduce the need for manual workflow boards, but strong growth can offset risk.

Forward 40Conviction 50Debt/Cash 21Evidence 42
READ THE RECONSTRUCTED THESIS ->

#22 · AS-OF October 18, 2024 · DUOL

Short DUOL (Duolingo)

Reconstructed as-of thesis: general-purpose AI tutors can compete with language-learning subscriptions, but brand, habit loops, and product velocity are strong offsets.

Forward 37Conviction 45Debt/Cash 20Evidence 42
READ THE RECONSTRUCTED THESIS ->

Turn the archive into a live watchlist.

The Daily Short stays separate from this backtest archive, but the same data model can feed alerts, newsletters, and founder review.