Research / Valuation and historical cases

Historical cases updated · Historical cases + current watchlist

Valuation conditions

Compare operating assumptions, valuation conditions and financing risks through cloud, cooling and optical communications companies.

This research is over30days old. Review new filings and prices before applying its assumptions.

VRT · Earnings performance and valuation sensitivity

The company is profitable; the gap between the base valuation scenario and the reference price is limited.

The base scenario needs $17.5bn CY2027 sales and a 23% GAAP-like operating margin: 25% sales growth from 2026 guidance midpoint, plus margin improvement. It is a research assumption, not company guidance.

Dated reference:USD 241.49 ·

Bear$126.7 to $127.5-47.5% to -47.2% vs reference
Base$283.5 to $284.317.4% to 17.7% vs reference
Bull$476.5 to $477.397.3% to 97.6% vs reference

Major downside risks:Delivery delays, weaker pricing or reversing customer advances can hurt profit and cash together. Acquiring UIG adds $1.45bn upfront cash plus up to $1.15bn contingent cash.

Follow-up review points:Q3 results expected October/November 2026 (date unconfirmed); UIG closing announced for Q4 2026. Check revenue against $3.65–3.85bn Q3 guidance and actual acquisition funding. Track reported earnings, acquisition progress and financing effects against the stated scenarios. Valuation multiples are research assumptions; the bull scenario alone does not establish investment value.

Assumptions, capital structure and sources

EBIT means operating profit after depreciation, amortization and stock compensation.

EV/EBIT after depreciation, amortization, rent and stock compensation. The dated price and mixed-date capital structure imply about 33.5× 2026 guided GAAP EBIT before UIG adjustments. Research multiples 20/28/35× stress contraction, normalization and a sustained premium. They are not observed peer medians.

Cash scenarios deduct UIG upfront, contingent reserves and $234.6m legacy acquisition consideration once. Debt stays $2,967.2m as an assumption; pension $76.5m is deducted separately. Preferred/NCI assumed absent. Shares rise to 395/398/402m, distinct from weighted EPS shares. The displayed range tests an additional $0–316.4m lease deduction because the reviewed Q2 report does not separately disclose the long-term finance/operating split. Current operating leases of $82.0m are disclosed in Note 8; the sensitivity range is not a verified bound on all lease obligations.

Actual outstanding Class A common
384.988173 million · 2026-07-27 Source ↗
Cash + short-term investments
$3,110.6 million · 2026-06-30 Source ↗
Debt principal
$2,967.2 million · 2026-06-30 Source ↗
UIG cash purchase price / maximum contingent cash
$1,450 / $1,150 million · 2026-09-02 Source ↗
Net pension liability, carried forward as assumption
$76.5 million · 2025-12-31 Source ↗

Bear · Research assumptions

Flat sales to the 2026 guidance midpoint despite the acquisition; 18% margin and multiple compression. Cash bridge (USDm): 3110.6 + H2 2026 owner cash 700 + 2027 owner cash 1317.5 - 1450 UIG - 0 earnout reserve - 234.6 legacy consideration -150 dividends -300 allocation = 2993.5. 2027 owner cash = (EBIT-150 interest)×75% + D&A 400 - capex 560.0 - working capital 300. This includes researcher reserves, not a cash forecast.

All amounts below are USD million; shares are million shares.

EBIT = 14,000 × 18% = 2,520
Enterprise value = 2,520 × 20 = 50,400
Common equity = 50,400 + 2,993.5 cash + 0 other assets − 2,967.2 debt − 0 preferred − 0 minority claims − 76.5 other claims = 50,349.8
Per share before additional uncertain claims = 50,349.8 ÷ 395 = $127.47

At this multiple and capital structure, matching the reference price requires the operating metric of 4,771.9 million USD.

Base · Research assumptions

25% sales growth, 23% margin versus 2026 GAAP guidance about21.1%; growth and leverage must both deliver. Cash bridge (USDm): 3110.6 + H2 2026 owner cash 900 + 2027 owner cash 2206.25 - 1450 UIG - 575 earnout reserve - 234.6 legacy consideration -150 dividends -300 allocation = 3507.25. 2027 owner cash = (EBIT-150 interest)×75% + D&A 500 - capex 700.0 - working capital 500. This includes researcher reserves, not a cash forecast.

All amounts below are USD million; shares are million shares.

EBIT = 17,500 × 23% = 4,025
Enterprise value = 4,025 × 28 = 112,700
Common equity = 112,700 + 3,507.3 cash + 0 other assets − 2,967.2 debt − 0 preferred − 0 minority claims − 76.5 other claims = 113,163.6
Per share before additional uncertain claims = 113,163.6 ÷ 398 = $284.33

At this multiple and capital structure, matching the reference price requires the operating metric of 3,416.1 million USD.

Bull · Research assumptions

50% growth and26% margin exceed current guidance; strongest execution plus preserved premium required. Cash bridge (USDm): 3110.6 + H2 2026 owner cash 1100 + 2027 owner cash 2887.5 - 1450 UIG - 1150 earnout reserve - 234.6 legacy consideration -150 dividends -300 allocation = 3813.5. 2027 owner cash = (EBIT-150 interest)×75% + D&A 650 - capex 945.0 - working capital 800. This includes researcher reserves, not a cash forecast.

All amounts below are USD million; shares are million shares.

EBIT = 21,000 × 26% = 5,460
Enterprise value = 5,460 × 35 = 191,100
Common equity = 191,100 + 3,813.5 cash + 0 other assets − 2,967.2 debt − 0 preferred − 0 minority claims − 76.5 other claims = 191,869.8
Per share before additional uncertain claims = 191,869.8 ÷ 402 = $477.29

At this multiple and capital structure, matching the reference price requires the operating metric of 2,751.7 million USD.

Same business outcome, different valuation multiple

Fix revenue, margin, cash, debt and shares at the base assumptions. These are valuation stress choices, not peer medians.

20× EBIT$202.6 to $203.4-16.1% to -15.8%
28× EBIT$283.5 to $284.317.4% to 17.7%
35× EBIT$354.3 to $355.146.7% to 47.1%

Evidence and model limitations:UIG terms disclose $1.45bn upfront cash at approximately 13× management-expected 2027 target EBITDA; this is not independently verified historical target earnings or a final purchase-price allocation. The model’s valuation multiples are scenario assumptions, not peer comparisons or consensus estimates. The Q2 report discloses $316.4m of long-term lease liabilities and $82.0m of current operating leases, but does not separately identify the long-term finance/operating split. The $76.5m pension deduction carries forward the December 2025 balance as an assumption. Future awards and acquisition outcomes remain scenario inputs.

COHR · Cash conversion and valuation assumptions

FY2026 operating cash flow of $79.5m was below $1.103bn in capital expenditure; valuation remains sensitive to cash conversion and the assumed multiple.

At a fixed28× EBIT, the central case is about$208 versus the September17 reference$295.98. It assumes $10.6bn CY2027 revenue and15% operating margin, about1.6× Q4 FY2026 annualized EBIT. These values depend on the stated operating and valuation assumptions.

Dated reference:USD 295.98 ·

Bear$110.2-62.8% vs reference
Base$208.1-29.7% vs reference
Bull$308.44.2% vs reference

Major downside risks:FY2026 operating cash was only $79.5m against $1.103bn capex. Inventory, yields or slower shipments can hurt cash and margins together; all three operating scenarios hold the multiple at 28× EBIT. Multiple risk is shown separately in the valuation sensitivity analysis.

Follow-up review points:Next quarterly results expected in calendar Q4 2026; exact earnings date unconfirmed. Compare sales with Q1 FY2027 guidance of $2.2–2.4bn. Track operating cash flow, capital expenditure and working-capital changes in subsequent reports. Compare price with explicitly stated valuation assumptions; the bull scenario alone does not establish investment value.

Assumptions, capital structure and sources

EBIT means operating profit after depreciation, amortization and stock compensation.

CY2027 income bridge: GAAP gross margin minus R&D, SG&A and other operating costs, retaining depreciation/amortization and SBC. Q4 FY2026 actual anchors:38.5% gross margin,10.6% R&D,13% SG&A,12.4% EBIT. All operating cases fix28× EBIT; the separate20/28/36× table changes only the multiple. A mixed-date $59.4bn EV proxy divided by the Q1 guidance GAAP-like annualization gives about45×;28× stresses about38% compression. The proxy subtracts the previous quarter$192m GAAP/adjusted gap from guidance—it is not management GAAP guidance, a peer multiple or a justified fair price.

Start with June 30 cash/investments of $1,987.018m and debt including finance leases of $3,271.715m; August 10 actual shares are a different date. Exclude $606.378m restricted cash. Preferred conversion and NVIDIA placement are already included. Note 15 does not provide a current minority fair-value estimate: the deductions below are explicit sensitivity assumptions. Operating rents remain in EBIT, so operating lease liabilities are not deducted again.

Actual common shares
195.832246 million · 2026-08-10 Source ↗
Cash + short-term investments
$1,987.018 million · 2026-06-30 Source ↗
Debt principal + finance leases
$3,271.715 million · 2026-06-30 Source ↗
Preferred stock remaining
$0; converted shares already included · 2026-06-30 Source ↗
Minority book value, not fair value
$334.705 million · 2026-06-30 Source ↗
Operating cash flow / capital expenditure
$79.514 / $1,102.909 million · FY2026 ended 2026-06-30 Source ↗
Outstanding options / unvested RSU units
376,536 / 2,406,714 · 2026-06-30 Source ↗

Bear · Research assumptions

Revenue below annualized Q1 midpoint9.2bn; GM37% deteriorates1.5pp from Q4, RD/SGA deleverage to11%/14%; remaining othercost1%. Cashconversion only partially recovers. EBIT margin=37% gross-11% R&D-14% SG&A-1% other. Other cost1% assumes easing from Q4~2.5%. July2026–December2027 independent OCF/capex/netdebt assumptions (USDm):1000/1800/500. Derived cash=1987.018+OCF-capex+netdebt=1687.018. Capex annualized remains at/above FY26$1103m; even bear OCF assumes recovery. No full working-capital forecast proves this funding feasible.

All amounts below are USD million; shares are million shares.

EBIT = 8,200 × 11% = 902
Enterprise value = 902 × 28 = 25,256
Common equity = 25,256 + 1,687 cash + 0 other assets − 3,771.7 debt − 0 preferred − 350 minority claims − 0 other claims = 22,821.3
Per share before additional uncertain claims = 22,821.3 ÷ 207 = $110.25

At this multiple and capital structure, matching the reference price requires the operating metric of 2,275.1 million USD.

Base · Research assumptions

Revenue15.2% above annualized Q1 midpoint9.2bn; GAAP GM39.5% improves1pp vs Q4, RD10.5%, SGA13%, othercost1%; EBIT15% is derived, not selected independently. EBIT margin=39.5% gross-10.5% R&D-13% SG&A-1% other. Other cost1% assumes easing from Q4~2.5%. July2026–December2027 independent OCF/capex/netdebt assumptions (USDm):2100/2400/0. Derived cash=1987.018+OCF-capex+netdebt=1687.018. Capex annualized remains at/above FY26$1103m; even bear OCF assumes recovery. No full working-capital forecast proves this funding feasible.

All amounts below are USD million; shares are million shares.

EBIT = 10,600 × 15% = 1,590
Enterprise value = 1,590 × 28 = 44,520
Common equity = 44,520 + 1,687 cash + 0 other assets − 3,271.7 debt − 0 preferred − 700 minority claims − 0 other claims = 42,235.3
Per share before additional uncertain claims = 42,235.3 ÷ 203 = $208.06

At this multiple and capital structure, matching the reference price requires the operating metric of 2,227.5 million USD.

Bull · Research assumptions

Revenue41.3% above annualized Q1 midpoint9.2bn; GAAP GM41.5% improves3pp vs Q4, RD10%, SGA12.5%, othercost1%; EBIT18% needs utilization/mix and is not guidance. EBIT margin=41.5% gross-10% R&D-12.5% SG&A-1% other. Other cost1% assumes easing from Q4~2.5%. July2026–December2027 independent OCF/capex/netdebt assumptions (USDm):3300/3000/-500. Derived cash=1987.018+OCF-capex+netdebt=1787.018. Capex annualized remains at/above FY26$1103m; even bear OCF assumes recovery. No full working-capital forecast proves this funding feasible.

All amounts below are USD million; shares are million shares.

EBIT = 13,000 × 18% = 2,340
Enterprise value = 2,340 × 28 = 65,520
Common equity = 65,520 + 1,787 cash + 0 other assets − 2,771.7 debt − 0 preferred − 1,000 minority claims − 0 other claims = 63,535.3
Per share before additional uncertain claims = 63,535.3 ÷ 206 = $308.42

At this multiple and capital structure, matching the reference price requires the operating metric of 2,248.4 million USD.

Same business outcome, different valuation multiple

Fix revenue, margin, cash, debt and shares at the base assumptions. These are valuation stress choices, not peer medians.

20× EBIT$145.4-50.9%
28× EBIT$208.1-29.7%
36× EBIT$270.7-8.5%

Evidence and model limitations:Valuation multiples are research scenarios, not a completed peer-valuation study or sourced consensus. FY2026 Note 15 reports NCI book value of $334.705m, without a current fair-value estimate; modeled deductions are sensitivities, not reported fair value. Note 17 discloses outstanding options and unvested awards. The August 31 Form 8-K adds performance awards with service and performance conditions through 2030. These are disclosed arrangements, not shares already issued; future vesting remains conditional. Note 14 classifies the NVIDIA price-protection provision as equity, not a derivative liability. The reviewed 10-K and subsequent August 31 8-K do not establish a later settlement result; potential cash or share settlement remains conditional, rather than a confirmed payment.

NBIS · Financing and delivery risk

The contribution of business growth to per-share value depends on financing costs, dilution and delivery progress.

H1 statements disclose $37.49bn remaining performance obligations, but only 36% is expected within 24 months to June 2028. This is not 2027 revenue or cash available to shareholders. A September 8 Palantir partnership adds a distribution route after integration, but its announcement gives no contracted revenue amount.

Major downside risks:New debt, pre-funded warrants, share issuance and GPU replacement can absorb business growth. Customer prepayments explain most H1 operating cash and cannot be treated as recurring free cash.

Follow-up review points:Next quarterly filing and every financing 6-K: revenue-producing capacity, capex, unrestricted cash, ATM issuance and conversion terms. Monitor commissioned capacity and revenue contribution, capital expenditure and cash consumption, actual ATM issuance, and convertible conversion or cash settlement. These are future operating and capital-structure outcomes.

Assumptions, capital structure and sources

Working-capital and funding test, not P/E: with a $2bn minimum cash buffer, cumulative July 2026–December 2027 OCF/capex assumptions of $12/30bn, $22/40bn and $34/50bn imply additional funding needs of $5.50bn, $5.50bn and $3.50bn before new financing. These are researcher stresses, not management guidance or secured funding.

The funding scenario assumes full drawdown of the July $775m facility and uses August estimated net proceeds of $5,680m. Added to June cash, this gives $14,497.1m before subsequent operating/investing flows; it is not reported cash available today. The reviewed H1 note confirms the facility agreement, not full drawdown. The $800m debt exchange adds shares, not cash. Restricted cash, pre-funded warrants and previously raised NVIDIA cash cannot be added twice. Both June notes have100m assumed remaining principal after exchange. Contractual accrued principal at December31,2027 is112.269399m +109.584153m; coupon is separate. This does not resolve all convertible claims.

Actual Class A + B common shares
271.855218 million · 2026-06-30 Source ↗
Pre-funded warrant shares, additional
21.065936 million · 2026-06-30 Source ↗
Unrestricted cash
$8,042.1 million · 2026-06-30 Source ↗
Debt carrying value
$8,545.7 million · 2026-06-30 Source ↗
August new notes / exchanged shares
$5,750 million / approximately 15.8 million shares · 2026-08-24 Source ↗
Five-year Class A issuance authorization; excludes Class C; not shares issued
≤20% · 2026-08-26 Source ↗

Evidence and model limitations:Reviewed H1 statements disclose 12,729,493 ATM shares sold through June 30. August closing documents disclose approximately 15.8m shares under debt-exchange agreements, but do not provide a precise consolidated share count after all subsequent transactions. The September Palantir release does not quantify contracted revenue. The June 2029/2031 contracts support principal accrual, while the H1 maturity table differs by $12.5m for each original tranche without explaining the presentation. This page presents funding stress scenarios, not an NBIS per-share target. Future conversion timing, settlement choices and option time value remain modeling assumptions.

Download the assumptions, calculations and screening record (JSON)

Coverage and screening record:69 directory companies

This is a source-availability triage of the existing US-listed directory, not a market-wide value screen. Three companies are research cases across cooling, optics and cloud; they are not preselected winners. BE/SNDK remain historical references. Other names provide business-role coverage only; this scope does not rank or exclude investments.

Recorded 2026-09-14 UTC; 30 of69 had a revenue source in the stored snapshot. This is not verification of all current financials.

This valuation research began on 2026-09-14 and has no historical performance record. VRT and COHR have conditional scenarios; NBIS focuses on funding needs and dilution risk.

NYSE: BE · Bloom Energy

Power demand and order execution

Distinguish agreement ceilings, firm orders and cash collection when assessing how demand translates into operating results.

BE · Latest reviewed developments · 2026-09-11

Revenue growth and improving cash flow

Q2 2026 revenue: $1.065bn · +165.5% YoY

Cash generation still relies heavily on customer advances; delivery must follow.

View analysis and sources

Q2 2026 revenue was $1,065.4m, up 165.5%; GAAP gross margin 33.4%; operating cash flow +$226.4m.

Evidence supporting the case

The 2025 cash-burn checkpoint has been followed by positive quarterly cash generation and GAAP operating income of $182.2m.

Risks and limitations

Deferred revenue/customer deposits contributed $211.7m to quarterly cash flow; receivables and contract assets absorbed cash. Track delivery against advances, not just cash inflow.

Revenue rose 165.5% from a year earlier. Operating cash flow was positive in the latest quarter, but customer advances account for a large part of the inflow.

Investment implications
Operating results support the assessment of growing power demand. Follow-up research should verify delivery and profitability of systems covered by customer advances.
Follow-up indicators
Equipment deliveries, cash collection and changes in profit.
Check original disclosure ↗
  1. 2026-08-19 · Installation product update

    Bloom introduced Power Connect, claiming over 40% shorter onsite installation time through factory-integrated electrical work.

    The time reduction is a company claim, not an independent fleet-wide result; verify project adoption and completed installations.

    Source ↗
  2. 2026-09-01 · Japan channel expansion

    Hitachi announced collaboration with Bloom for onsite power at data centers and industrial facilities in Japan, following an Omika Works pilot.

    The release gives no firm order value or MW commitment; collaboration and a pilot are not booked sales.

    Source ↗
Revisit four historical checkpoints
  1. · Customer validation

    AEP agreement: up to 1 GW, with an initial 100 MW order.

    A named customer gives the power-shortage thesis a concrete beneficiary.

    The 1 GW ceiling is not 1 GW of firm orders or revenue.

    Original disclosure ↗
  2. · Financial confirmation

    FY2024 revenue: $1,473.9m, up 10.5%; operating cash flow: +$92.0m.

    The operating evidence extends beyond announcements.

    Common shareholders still had a $29.2m GAAP net loss. Positive operating cash flow does not imply positive P/E.

    Original disclosure ↗
  3. · Hyperscaler use case

    Oracle collaboration targeted onsite power delivery within 90 days at selected U.S. data centers.

    Speed to power becomes a specific customer need.

    This was a delivery target; the announcement did not disclose MW or contract dollars.

    Original disclosure ↗
  4. · Growth, with a cash test

    Q2 revenue: $401.2m, up 19.5%; GAAP gross margin: 26.7%.

    Follow deliveries and collections as manufacturing expands.

    Quarterly operating cash outflow: $213.111m. The 1-to-2 GW capacity increase was a plan for end-2026.

    Original disclosure ↗
Valuation lens
A loss-making GAAP earnings base cannot support a normal positive P/E. Test revenue conversion, sustainable margins, funding needs and dilution before assigning a valuation.
Conditions that invalidate the investment thesis
Repeated delivery delays, weak collections or dilution that overwhelms per-share growth would weaken the case. One seasonal cash-flow quarter alone is not a verdict.

NASDAQ: SNDK · Sandisk

Pricing and earnings recovery in the storage cycle

Assess recovery through pricing, gross margins and customer qualification. Sequential improvement does not imply that year-over-year earnings have recovered.

SNDK · Latest reviewed developments · 2026-09-11

The recovery has become strong reported profit

FY2026 Q4 revenue: $8.965bn · quarter ended July 3

Strong pricing may not last; do not annualize one exceptional quarter.

View analysis and sources

FY2026 Q4 ended July 3: revenue $8.965bn; GAAP gross margin 84.6%; GAAP net income $6.903bn. Datacenter revenue $2.977bn, +103% sequentially.

Evidence supporting the case

Management attributes approximately two-thirds of sequential revenue growth to pricing; test whether contract economics sustain it.

Risks and limitations

FY2027 Q1 revenue $10.30–10.80bn is guidance, not achieved revenue. Do not annualize one exceptional quarter.

The latest quarter shows much higher revenue and profit. The important question now is how long strong pricing can last.

Investment implications
Sales growth is reaching profit. Still, a strong business can be expensive if the share price already assumes years of unusually high margins.
Follow-up indicators
Contract delivery, product pricing, operating cash flow and capital expenditure.
Check original disclosure ↗
  1. 2026-08-12 · Technology roadmap

    Kioxia and Sandisk unveiled ninth-generation 2Tb QLC NAND; claimed interface speed is 4.8Gb/s, 33% above the preceding generation.

    Speed reflects the issuer’s test conditions, not whole-system throughput. A technology unveiling does not establish customer revenue.

    Source ↗
  2. 2026-08-13 · Contract visibility, forward model

    Investor Day: eight customers had signed NBM agreements; management estimates coverage of about 50% of FY2027 bits and two-thirds of FY2028 bits.

    FY2028–2030 targets of about 80% non-GAAP gross margin and 50% adjusted free-cash-flow margin are management expectations, not current results or guaranteed returns.

    Source ↗
Revisit four historical checkpoints
  1. · New starting point

    Sandisk began trading independently after separating from Western Digital.

    Use the new security’s history; do not splice in the former Sandisk listing.

    Separation itself does not establish cheap valuation.

    Original disclosure ↗
  2. · Early recovery signal

    FY2025 Q3 revenue: $1.695bn, down 10% sequentially. Management reported supply cuts and price increases.

    Look for pricing actions to reach reported margins in later quarters.

    GAAP loss was $1.933bn, including $1.83bn goodwill impairment; this is not all cash burn.

    Original disclosure ↗
  3. · Operating improvement

    FY2025 Q4 revenue: $1.901bn; GAAP gross margin: 26.2%, versus 22.5% the prior quarter.

    Revenue and margin improved together, a stronger signal than a product announcement.

    GAAP net loss remained $23m. The quarter was not yet GAAP net profitable.

    Original disclosure ↗
  4. · Profit confirmation

    FY2026 Q1 revenue: $2.308bn; GAAP net income: $112m. Datacenter revenue rose 26% sequentially.

    Customer qualification can help test whether recovery has a durable demand engine.

    Datacenter revenue still fell 10% year on year. Qualification is not a delivered order.

    Original disclosure ↗
Valuation lens
Losses and impairment distort trailing P/E. Use normalized cycle earnings, cash generation and the share count known at each date; do not annualize one strong quarter.
Conditions that invalidate the investment thesis
A renewed pricing decline, margin reversal or stalled customer qualification would weaken the recovery thesis.
Comparative analysis: business and financing differences

PLUG is a check against an overly broad fuel-cell screen, not a matched BE peer or a proven failed-return example. Its 2024 operating cash use was $728.6m. The research question is whether unit economics and financing can support growth.

PLUG 2024 10-K · Liquidity and Capital Resources ↗

To validate a selection rule, keep all qualifying companies—including disappointments—and test only information available before each decision. Two selected winners cannot establish a hit rate.

How earnings growth and valuation affect share prices

For positive earnings, share-price growth equals EPS growth × the change in P/E. If the valuation multiple halves, earnings per share must grow twice as much to reach the same price target.

Illustrative 5-year hurdles · not forecasts
Price multipleAnnualized price growthEPS multiple if P/E halves
5×38.0%10×
10×58.5%20×
100×151.2%200×
Investment research framework
  1. Pressure: identify a measurable constraint, not just more news coverage.
  2. Beneficiary: link the constraint to a named customer, qualified product or firm order.
  3. Economics: follow delivery, margins, cash collection and dilution.
  4. Price: compare bear, base and bull assumptions with a dated price and share count; test the implied market size.
  5. Accountability: preserve the original call and record evidence that strengthens or weakens it.

Historical entry returns and drawdowns require a consistent, corporate-action-adjusted price series. These cases focus on operating evidence and do not include entry-return or drawdown comparisons. The framework has not been validated as a predictive score.