Research note · Cross-sectional analysis

ETF Universe · Exposure, Skill, and Why SPY Is So Hard to Beat

Name: ETF universe evidence · Study 1A
Creator: Carlos Barredo Lago
Published: 2026-08-31

A stepwise test of the free-data universe: describe it first, separate exposure from management next, then ask whether the result survives cash, financing and SPY as the investor's opportunity cost.

QNV-20260831-01 31 August 2026 Yahoo Adj Close · FRED DFF Study 1A

GitHub Code and reproduction

Executive thesis

1 Product abundance is not skill abundance

Traditional metrics remain in place to describe the 2,023 mature ETFs. To decide which products produced a superior path to SPY, we avoid a vote among ratios: RWM is the primary criterion and CAGR provides economic context. 161 products (8.0%) beat SPY's RWM over the same window; 93 (4.6%) also beat its CAGR.

What RWM measures: Relative-Wealth Martin calculates the Martin ratio entirely on ETF wealth relative to cash. It relates the annualised growth of that advantage to the Ulcer Index of the same curve: gaining ground versus cash is rewarded, while the depth and persistence of losing it are penalised. Here it condenses relative growth and damage into one reading; the metric itself is not the subject of this report. Read more about RWM at Qinvia Research ↗

161beat RWM · 8.0%

93also beat CAGR

95identifiable beta exposure

15intentional management

The conclusion up front: a semiconductor, gold or technology index winning during the sample identifies a rewarded exposure. It does not demonstrate manager selection, timing or risk control. Commercial scale, complexity and brand recognition are not evidence of alpha either.

The initial scorecard, metric by metric

Criterion	Beat SPY	Do not beat SPY	% of universe
CAGR	213	1,810	10.5%
Sortino	201	1,822	9.9%
Calmar	242	1,781	12.0%
Martin	217	1,806	10.7%
RWM	161	1,862	8.0%
CAGR + RWM	93	1,930	4.6%
All four at once	59	1,964	2.9%

CAGR, Sortino, Calmar and Martin remain descriptive. From here on, RWM drives the comparison and CAGR is the second coordinate—not another vote.

Universe and method

2 A demanding comparison that remains reproducible—and honest about its limits

The raw catalogue preserves 9,550 identities, including historical records. Inverse and leveraged exposures, volatility products, structured payoffs and vehicles without verified trading are set aside, not deleted. The analytical sample requires a usable Yahoo start on or before 1 March 2022 and at least 1,008 daily returns.

Same window, total return

Each ETF is aligned with SPY from its first usable session to its own final session. Yahoo Adj Close is used as a total-return proxy, preventing different periods from being compared as if they were equivalent.

SPY as opportunity cost

The question is deliberately hard: “would holding SPY have been better?”. We do not claim that SPY is the natural benchmark for bonds, gold or market-neutral mandates; it is the bar faced by a growth-seeking investor.

Operating rule: the tables retain CAGR, Sortino, Calmar and Martin for a complete diagnostic. Superiority is decided by comparing each ETF's RWM with SPY's over identical sessions; CAGR remains beside it to separate efficiency from absolute growth.

The unit changes from SEC identity to Yahoo symbol and then to economic product. Ticker changes are consolidated to avoid double counting.

Survivorship bias remains material: 1,956 of 2,023 products are flagged as current and only 67 (3.3%) as non-current. Moreover, all 67 non-current series extend at least to 17 July 2026; they are not a representative history of dead ETFs. The free price cohort therefore cannot support a claim that survivorship bias has been materially removed.

Main exclusion groups

Summary reason	Products
Leveraged or inverse	1358
Structured payoff or tail protection	538
Non-tradable symbol or vehicle	444
Never launched or no verified trading	53
Volatility products	8

Maturity cutoff: 1 March 2022 requires every product to have lived through at least the selected inflation shock and roughly four and a half years of market history by the data close. CGDV and CGXU qualify: both have a first Yahoo session of 24 February 2022.

The cutoff does not equalise age. Each ETF is still measured from its own inception; temporal robustness is examined later.

Identities, aliases and unresolved classification

2,028 series pass the initial filter. Five are former tickers for the same product and are consolidated, leaving 2,023 economic products.

Previous ticker	Current ticker	Product
EWRE	RSPR	Invesco S&P 500 Equal Weight Real Estate ETF
GLDX	USG	USCF Gold Strategy Plus Income Fund
GRZZ	DARP	Grizzle Growth ETF
HSPX	XYLD	Global X S&P 500 Covered Call ETF
SDEF	FXED	Sound Enhanced Fixed Income ETF

Primary class still unresolved

Ticker	Product	Evidence
GTR	WisdomTree Target Range Fund	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic
IPOS	Renaissance Capital Greenwich Fund	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic
QQQ	Invesco QQQ Trust, Series 1	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic
SCCD	Sachem Capital Corp. 6.00% Note	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic
SWZ	Total Return Securities, Inc.	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic
WDNA	WisdomTree BioRevolution Fund	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic
ZIG	The Acquirers Fund	asset:unknown\|exposure:unclassified\|management:free_metadata_insufficient\|geography:not_determined\|theme:not_thematic

CAGR = (V_T/V₀)^1/years − 1 · Sortino = mean(r)/downside deviation × √252

Calmar = CAGR/|MaxDD| · Martin = CAGR/Ulcer Index

Exposure is not skill

3 Before assigning credit, identify the risk that was bought

The taxonomy separates intentional management from broad-market beta, concentrated beta, static systematic rules, fixed income and other exposures. The aim is not to diminish index investing; it is to stop a winning sector or asset from being presented as evidence of superior management.

Origin of products that beat RWM and those that also beat CAGR. “Unresolved” stays explicit rather than being forced into active or passive.

Broad beta

A diversified exposure can be an excellent, inexpensive solution. Its success belongs to the market captured, however—not to a manager repeatedly making superior decisions.

Concentrated beta

Sectors, themes, countries, metals and styles narrow the portfolio and increase regime dependence. Winning ex post may reflect a premium or an exceptional period.

Systematic rules

An alternative index may embody a valid idea. Skill requires evidence that the rule adds value robustly, not merely that its backtest selected a winning exposure.

A useful example: QQQ beats SPY's CAGR since 1999 (10.8% versus 8.6%), but not its RWM (0.171 versus 0.295). A narrower index may capture an extraordinary regime without delivering a superior relative path under the chosen criterion.

The central test

4 What remains when the product is expected to make decisions?

Two layers prevent intention from being confused with skill: 128 products with identified active management and a wider ring of 11 dynamic or alternative strategies whose active label is not confirmed. The extended block contains 139 products; economic exposure remains a separate axis.

Results for the extended active/dynamic block: CAGR as context, RWM as the criterion and their intersection as the stricter reading.

15beat RWM · 10.8%

9beat RWM + CAGR

13identified active · RWM

2inferred dynamic · RWM

Observed proportions and 95% Wilson intervals. Their overlap is a reminder that a management label is not causal evidence of added value.

The bar stays high: 13 of 128 identified active products (10.2%) beat SPY's RWM; 9 also beat its CAGR. Across the extended ring, the total rises to 15 of 139. A process has been identified; that does not prove the process caused the outcome.

The 15 cases that beat SPY's RWM

Ticker	Product	Strategy	CAGR	SPY CAGR	RWM	SPY RWM	Beats CAGR
CLSE	Convergence Long/Short Equity ETF	Long/short and market neutral	20.3%	15.4%	2.629	1.298	True
CGDV	Capital Group Dividend Value ETF	Dividend / income	19.1%	15.5%	2.310	1.307	True
AVLV	Avantis U.S. Large Cap Value ETF	Factor / style	14.9%	13.4%	1.694	0.963	True
SIXH	ETC 6 Meridian Hedged Equity Index Option ETF	Long/short and market neutral	11.4%	18.3%	2.418	1.744	False
DFIV	Dimensional International Value ETF	Factor / style	16.1%	13.2%	1.590	0.950	True
TEQI	T. Rowe Price Equity Income ETF	Dividend / income	15.4%	16.5%	1.913	1.499	False
JAVA	JPMorgan Active Value ETF	Factor / style	12.5%	14.0%	1.390	1.021	False
HEQT	Simplify Hedged Equity ETF	Long/short and market neutral	9.4%	12.7%	1.084	0.878	False
AVIV	Avantis International Large Cap Value ETF	Factor / style	14.4%	14.2%	1.186	1.041	True
CGUS	Capital Group Core Equity ETF	Equity · other	15.8%	15.5%	1.437	1.307	True
DUHP	Dimensional US High Profitability ETF	Equity · other	14.0%	15.5%	1.434	1.307	False
TSPA	T. Rowe Price U.S. Equity Research ETF	Equity · other	14.1%	13.7%	1.100	1.043	True
AVUS	Avantis U.S. Equity ETF	Equity · other	16.4%	16.4%	1.532	1.477	True
DFAU	Dimensional US Core Equity Market ETF	Equity · other	15.6%	15.8%	1.385	1.376	False
THRO	iShares U.S. Thematic Rotation Active ETF	Thematic equity	13.4%	12.8%	0.873	0.867	True

Lowering the return bar—without giving away the result

5 6 candidates show better RWM but still trail SPY's return

We select 6 products from the active/dynamic block that do not beat SPY's CAGR but do beat its RWM: 4 have identified active management and 2 are inferred dynamic strategies. This is the right place to ask whether a more efficient relative path can become competitive through moderate leverage.

Why they qualify

RWM recognises growth over cash and penalises periods in which relative wealth remains below prior highs. It identifies useful paths without prematurely declaring absolute winners.

What they have not earned

Better RWM alone is not enough. They must match SPY's terminal wealth after financing the debt—and the RWM advantage must survive.

The 6 relative-efficiency candidates

Ticker	Product	Asset class	Strategy	Identified process	CAGR	SPY CAGR	RWM	SPY RWM
SIXH	ETC 6 Meridian Hedged Equity Index Option ETF	Alternatives	Long/short and market neutral	Identified index/passive	11.4%	18.3%	2.418	1.744
TEQI	T. Rowe Price Equity Income ETF	Equity	Dividend / income	Identified active	15.4%	16.5%	1.913	1.499
JAVA	JPMorgan Active Value ETF	Equity	Factor / style	Identified active	12.5%	14.0%	1.390	1.021
HEQT	Simplify Hedged Equity ETF	Alternatives	Long/short and market neutral	Not determined	9.4%	12.7%	1.084	0.878
DUHP	Dimensional US High Profitability ETF	Equity	Equity · other	Identified active	14.0%	15.5%	1.434	1.307
DFAU	Dimensional US Core Equity Market ETF	Equity	Equity · other	Identified active	15.6%	15.8%	1.385	1.376

Leverage ladder

6 Leverage can lift CAGR. It cannot manufacture skill.

Exposure is purchased once with fixed initial debt and then allowed to evolve. There is no daily rebalancing or invented turnover: leverage drifts with the market. Debt accrues the daily Effective Federal Funds Rate plus a 1.50% base spread under Actual/360.

Equity_t = L × ETF_t − (L−1) × Debt_t

Counts refer to the 6 candidates. Leverage is fixed at inception; constant exposure is not targeted.

Initial leverage	Candidates	Beat CAGR	Preserve RWM	Beat both
1.00×	6	0	6	0
1.25×	6	4	5	3
1.50×	6	4	5	3
1.75×	6	4	3	2
2.00×	6	5	3	2

Reading the ladder: at 1.25×, four products already beat SPY's CAGR, five preserve RWM and three achieve both. At 2.00×, five reach the CAGR, but only three retain RWM and two retain both criteria. Leverage does not create skill; it exchanges path headroom for growth.

The financing structure is a historical “IBKR-style” scenario, not an exact reconstruction of every past rate. IBKR publishes rates as benchmark plus spread, calculated daily and, for USD, on Actual/360. Official sources: margin rates, calculation method and the Effective Federal Funds Rate (FRED).

Exact return matching

7 Same ending. Very different journey.

We solve ex post for the exact initial leverage required to finish with the same wealth as SPY. This is a diagnostic—not an invertible rule—and it is capped at 2.00×. 5 of 6 candidates can reach the target; 4 retain a higher RWM after financing.

Different paths to the same terminal wealth

Each panel begins at 100 on the ETF's inception date. The levered line includes historical financing and ends beside SPY by construction; the relevant difference is the path of relative wealth.

Return matching and financing sensitivity

Left: RWM after matching return. Right: sensitivity to the financing spread over Fed Funds.

5reach SPY at ≤ 2×

4preserve higher RWM

1.50%base spread over DFF

2.00×diagnostic cap

Outcome by product

Ticker	Product	Strategy	Identified process	Required leverage	Matched RWM	SPY RWM	Outcome
DFAU	Dimensional US Core Equity Market ETF	Equity · other	Identified active	1.03×	1.381	1.376	Matches and preserves RWM
TEQI	T. Rowe Price Equity Income ETF	Dividend / income	Identified active	1.14×	1.904	1.499	Matches and preserves RWM
DUHP	Dimensional US High Profitability ETF	Equity · other	Identified active	1.21×	1.372	1.307	Matches and preserves RWM
JAVA	JPMorgan Active Value ETF	Factor / style	Identified active	1.24×	1.315	1.021	Matches and preserves RWM
HEQT	Simplify Hedged Equity ETF	Long/short and market neutral	Not determined	1.97×	0.843	0.878	Matches; loses RWM
SIXH	ETC 6 Meridian Hedged Equity Index Option ETF	Long/short and market neutral	Identified index/passive	2.42×	2.292	1.744	Requires more than 2×

Financing-cost sensitivity

Spread over Fed Funds	Feasible at ≤ 2×	Preserve RWM
1.0%	5	5
1.5%	5	4
2.0%	4	4
2.5%	4	3

Central result: DFAU, DUHP, JAVA and TEQI match terminal wealth below 2× and preserve higher RWM. HEQT reaches the return but loses its RWM advantage; SIXH would require more than 2×. The conclusion deteriorates as debt becomes more expensive.

The 25% maintenance threshold is illustrative; actual requirements depend on broker, portfolio and product. Return matching uses ex-post information and must not be interpreted as an executable strategy.

Entry-point stability

8 A launch-to-date win is not enough

The 15 RWM winners in the managed block are compared with SPY again from different entry months using rolling windows of roughly three years. 7 beat SPY's RWM from a majority of entry points. Because the windows overlap, this is temporal sensitivity—not an independent estimate of persistence.

Fraction and share of entry months whose subsequent 756-session window beats SPY's RWM; CAGR remains in the table as context.

Ticker	Product	Beats RWM	% CAGR	% RWM	% CAGR + RWM	Worst Δ CAGR	Worst Δ RWM
CGDV	Capital Group Dividend Value ETF	18/18	100.0%	100.0%	100.0%	1.9%	0.670
CGUS	Capital Group Core Equity ETF	18/18	94.4%	100.0%	94.4%	−0.2%	0.108
TSPA	T. Rowe Price U.S. Equity Research ETF	26/27	100.0%	96.3%	96.3%	0.3%	−0.071
CLSE	Convergence Long/Short Equity ETF	16/19	84.2%	84.2%	78.9%	−0.5%	−0.367
DFIV	Dimensional International Value ETF	17/24	62.5%	70.8%	62.5%	−5.1%	−0.483
SIXH	ETC 6 Meridian Hedged Equity Index Option ETF	24/40	10.0%	60.0%	10.0%	−13.8%	−3.129
THRO	iShares U.S. Thematic Rotation Active ETF	11/21	90.5%	52.4%	52.4%	−0.6%	−0.554
HEQT	Simplify Hedged Equity ETF	10/22	4.5%	45.5%	4.5%	−9.0%	−1.003
AVUS	Avantis U.S. Equity ETF	21/47	38.3%	44.7%	34.0%	−3.7%	−1.494
AVIV	Avantis International Large Cap Value ETF	9/23	30.4%	39.1%	30.4%	−6.4%	−0.646
TEQI	T. Rowe Price Equity Income ETF	13/37	13.5%	35.1%	13.5%	−11.7%	−3.186
AVLV	Avantis U.S. Large Cap Value ETF	7/23	21.7%	30.4%	8.7%	−7.1%	−2.406
DUHP	Dimensional US High Profitability ETF	5/18	11.1%	27.8%	11.1%	−4.9%	−1.195
JAVA	JPMorgan Active Value ETF	4/23	4.3%	17.4%	4.3%	−10.0%	−2.883
DFAU	Dimensional US Core Equity Market ETF	0/34	0.0%	0.0%	0.0%	−1.4%	−0.797

An important separation: CGDV, CGUS, TSPA, CLSE, DFIV, SIXH and THRO beat RWM from at least half of the entry months. The others depend much more heavily on the chosen start. Apparent percentage precision should not be mistaken for independent samples.

Synthesis and contrast

9 One decision criterion, one complementary descriptor

RWM orders the comparison with SPY; CAGR keeps economic magnitude visible. DBF± does not select winners in this study: it describes whether net return direction is supported by breadth.

The 161 points above zero on RWM beat SPY under the primary criterion; 93 also lie to the right of zero on CAGR. Visual limits trim extremes, but counts use every observation.

161 beat RWM

8.0% of the universe has a higher RWM than SPY over its comparable window.

93 beat RWM + CAGR

4.6% preserves the relative advantage and also finishes with higher annualised growth.

The fifteen largest RWM advantages

Ticker	Product	Process	CAGR	SPY CAGR	RWM	SPY RWM	Δ RWM	Beats CAGR
FFLC	Fidelity Fundamental Large Cap Core ETF	Systematic / rules-based	19.7%	16.8%	3.432	1.566	1.866	True
PSMJ	Pacer Swan SOS Moderate (July) ETF	Not determined	10.5%	13.3%	2.623	0.995	1.628	False
BALT	Innovator Defined Wealth Shield ETF	Not determined	6.0%	13.4%	2.597	1.012	1.585	False
PVAL	Putnam Focused Large Cap Value ETF	Systematic / rules-based	17.0%	13.8%	2.594	1.070	1.524	True
PSMO	Pacer Swan SOS Moderate (October) ETF	Not determined	10.4%	14.2%	2.480	1.046	1.434	False
RISR	FolioBeyond Alternative Income and Interest Rate Hedge ETF	Not determined	14.0%	13.9%	2.401	1.014	1.387	True
PSFO	Pacer Swan SOS Flex (October) ETF	Not determined	11.7%	13.9%	2.379	1.014	1.365	False
PSFJ	Pacer Swan SOS Flex (July) ETF	Not determined	11.4%	13.4%	2.361	1.012	1.349	False
CLSE	Convergence Long/Short Equity ETF	Identified active	20.3%	15.4%	2.629	1.298	1.331	True
PSFD	Pacer Swan SOS Flex (January) ETF	Not determined	12.8%	15.4%	2.339	1.319	1.020	False
CGDV	Capital Group Dividend Value ETF	Identified active	19.1%	15.5%	2.310	1.307	1.003	True
NETZ	Engine No. 1 Transform Climate ETF	Not determined	20.1%	13.1%	1.909	1.019	0.890	True
FLV	American Century Focused Large Cap Value ETF	Systematic / rules-based	16.4%	20.7%	2.830	2.049	0.782	False
PSFF	Pacer Swan SOS Fund of Funds ETF	Not determined	10.1%	15.2%	2.074	1.295	0.779	False
AVLV	Avantis U.S. Large Cap Value ETF	Identified active	14.9%	13.4%	1.694	0.963	0.731	True

D± = Σr_t / Σ|r_t| · J = (Σ|r_t|)² / (N·Σr_t²) · DBF± = D± × J

DBF is a contrast, not a verdict: it is invariant to order and positive scale. It does not observe drawdowns, economic severity, tail risk, alpha or future return; it is therefore not used to declare superiority to SPY.

The ten highest DBF± profiles

Ticker	Product	CAGR	DBF±	D±	J
SGOV	iShares 0-3 Month Treasury Bond ETF	3.0%	0.436	0.884	0.493
BILS	State Street SPDR Bloomberg 3-12 Month T-Bill ETF	2.9%	0.364	0.770	0.473
BKUI	BNY Mellon Ultra Short Income ETF	3.6%	0.265	0.478	0.555
VUSB	Vanguard Ultra-Short Bond ETF	3.4%	0.180	0.342	0.526
SHV	iShares 0-1 Year Treasury Bond ETF	1.6%	0.170	0.446	0.380
TBUX	T. Rowe Price Ultra Short-Term Bond ETF	4.0%	0.166	0.315	0.526
GBIL	Goldman Sachs Access Treasury 0-1 Year ETF	2.3%	0.153	0.576	0.266
GSST	Goldman Sachs Ultra Short Bond ETF	3.3%	0.142	0.384	0.370
CLTL	Invesco Treasury Collateral ETF	2.4%	0.128	0.540	0.236
EMNT	PIMCO Enhanced Short Maturity Active ESG Exchange-Traded Fund	3.1%	0.127	0.386	0.329

The universe as a whole

10 SPY does not merely beat the average—it finishes high in the distribution

Every ETF is rebased to 100 on 1 March 2022. The fan displays the full dispersion of buy-and-hold wealth; the second figure separates mean, median and coverage sensitivity. Neither represents a rebalanced portfolio.

Cross-sectional bands for 1,967 products with complete coverage. The fan shows the centre, dispersion and extremes at the same time.

The mean tells the same story

The complete-coverage sensitivity requires at least 98.5% observable sessions through the end.

190.6SPY terminal wealth

153.5mean ETF terminal wealth

15.4%SPY CAGR

10.0%mean ETF CAGR

Descriptive result: the mean curve is neither tradable nor rebalanced. It only summarises where the universe ended and how much dispersion surrounded that outcome.

Qinvia verdict

11 Simple does not mean easy to beat

What the evidence does show

Only 8.0% of the universe beats SPY's RWM.
Only 4.6% beats RWM and CAGR simultaneously.
Most attributable winners reflect beta exposure, not demonstrated management skill.
15 of 139 intentional-management products beat RWM; 7 do so from a majority of entry points.
After return matching and financing, 4 of 5 feasible cases preserve RWM.

Why the bar is structurally high

The S&P 500 already diversifies hundreds of profitable, liquid businesses.
Capitalisation weighting lets winners grow and mechanically reduces the weight of losers.
Its membership renews; it is not a static portfolio of the companies selected in 1957.
SPY packages that exposure with deep liquidity, low cost and little operational friction.
A more complex product therefore carries the burden of showing what improves—and whether the improvement persists.

Capacity, scale and the success paradox — hypothesis, not causal finding:

One explanation consistent with the literature is that alpha has finite capacity. Strong results attract assets under management; deploying larger positions across a limited opportunity set can increase market impact, costs, liquidity constraints and competition, eroding the edge that attracted the flows. Berk and Green (2004) formalise this equilibrium; Chen et al. (2004) find a more adverse size relationship among funds exposed to small and illiquid stocks, while Pástor, Stambaugh and Taylor find strong decreasing returns at the industry level but less conclusive evidence at the individual-fund level.

The counterpoint matters. Using real institutional trades, Frazzini, Israel and Moskowitz estimate far greater capacity than previously assumed for some strategies and show that cost-aware execution can expand it substantially. Capacity depends on turnover, horizon, liquidity, market breadth and execution. Our ETF data do not causally identify the effect of flows or estimate the capacity of each process; this mechanism is therefore presented as a plausible explanation, not a demonstrated conclusion.

Conclusion: the industry offers thousands of narratives, wrappers and exposures; evidence of management-added value is far scarcer. The result does not prove that the industry is a fraud or that every narrow exposure is useless. It does show that complexity, brand and assets under management are not enough: against a broad, inexpensive, self-renewing index, the burden of proof is exceptionally high.

Quality control and limits of free data

We validated 5,505 Parquet files without integrity failures. The catalogue preserves historical identities, but the 67 non-current series all reach recent dates and do not represent liquidations distributed through time. Survivorship bias remains a material limitation. RWM uses DFF as idealised institutional cash—daily accrual, no spread, Actual/360—and does not represent a guaranteed retail return. The taxonomy uses auditable evidence and leaves unresolved cases explicit.

Generated 2026-08-31 03:51 UTC. Primary source: free Yahoo Finance · daily Adj Close · exact dates shared with SPY. Cash: FRED DFF.

GitHub

Methodology, notebooks and reproducible materials.

The public repository preserves the bilingual notebooks, methodology and derived materials. This web publication retains the complete figures, formulas, charts, tables, limitations and conclusion.

GitHub Code and reproduction