dhawalorg
RECIPESReproducible analysisOpen JSON endpoints · local tools · disclosed methods

Show the inputs, transformation, and limits.

Each recipe starts with a research question and ends with an inspectable output. Calls are copyable, missing data stays missing, and the limitations travel with the method.

Recipes
6
Hidden inputs
None
Machine index
JSON

These are deterministic research paths through existing desk data. They do not add predictions or recommend trades. Endpoint responses retain their own source and freshness fields, which should be carried into any derivative output.

Recipe 01

Cross-asset tape check

Which liquid benchmarks are moving together during this session?

Inputs

  • Batched quotesLast price, session change, source, and observation time for each symbol.

Calls

curl "https://dhawal.org/api/quotes?symbols=SPY,QQQ,TLT,GLD,DX-Y.NYB,CL%3DF"

Transformation

  1. Retain only returned symbols and preserve each quote source and fetched-at value.
  2. Rank by session percentage change. Do not substitute a missing quote with zero.
  3. Compare direction across equities, duration, gold, the dollar, and crude oil.

Output

A timestamped cross-asset direction table with missing observations left explicit.

Limitations

  • Quotes may be delayed and may not share the same observation timestamp.
  • A one-session move does not establish a durable correlation or causal relationship.
Open productRead methodology

Recipe 02

Risk dial rebuild

Can the desk risk score be reconstructed from its disclosed legs?

Inputs

Calls

curl "https://dhawal.org/api/breadth"
curl "https://dhawal.org/api/options?symbol=SPY"
curl "https://dhawal.org/api/quotes?symbols=%5EVIX,%5EVIX9D,SPY,TLT"
curl "https://dhawal.org/api/macro"

Transformation

  1. Map each available leg to a zero-to-100 score using the formulas printed on Today.
  2. Clamp every leg to the zero-to-100 range.
  3. Take the equal-weight mean of non-missing legs and round to one decimal.
  4. Return no composite when fewer than three legs report.

Output

A score, zone, available-leg count, and the input detail for every reporting leg.

Limitations

  • Legs can mix market-session and economic-series observation dates.
  • The composite is descriptive. It is not a forecast or trading signal.
Open productRead methodology

Recipe 03

Yield-curve slope check

How steep or inverted is the current Treasury curve?

Inputs

Calls

curl "https://dhawal.org/api/yield-curve"

Transformation

  1. Read the latest available 2-year and 10-year nominal yields from the same response.
  2. Compute 10-year minus 2-year in percentage points.
  3. Multiply by 100 to express the slope in basis points.
  4. Carry the source date and any stale state into the output.

Output

A 2s10s slope in basis points with both input yields and their source date.

Limitations

  • Treasury constant-maturity yields are reference estimates, not executable prices.
  • Missing tenors must remain missing. Do not interpolate them for this recipe.
Open productRead methodology

Recipe 04

Seasonality cross-section

Which covered symbols have the strongest retained history for a calendar month?

Inputs

Calls

curl "https://dhawal.org/data/seasonality-index.json"

Transformation

  1. Choose one calendar month and a minimum history threshold before ranking.
  2. Exclude rows with fewer observations than the selected threshold.
  3. Rank by median monthly return, then inspect win rate and sample count together.
  4. Retain the generated-at timestamp from the index.

Output

A month-specific table with median return, win rate, sample size, and build timestamp.

Limitations

  • Seasonality describes historical samples and does not identify a catalyst.
  • Symbol histories can differ in length and survivorship context.
Open productRead methodology

Recipe 05

13F position-change audit

Which reported positions changed between a manager’s two retained filings?

Inputs

  • Normalized 13F bookCurrent holdings, prior-quarter changes, filing links, and reconciliation quality.

Calls

curl "https://dhawal.org/api/whales?filer=berkshire-hathaway"

Transformation

  1. Confirm that the current book is verified and inspect its reconciliation metadata.
  2. Separate new, exited, increased, decreased, and unchanged positions.
  3. Rank changes by absolute reported value change while preserving share change.
  4. Open the linked SEC filing before using a holding in downstream research.

Output

A filing-linked change ledger with quarter, filed date, value delta, share delta, and quality state.

Limitations

  • 13F reports are delayed, long-only snapshots and omit many asset and short exposures.
  • Issuer-to-ticker mapping can be incomplete, and amendments can revise a reported book.
Open productRead methodology

Recipe 06

Local portfolio shock

How would a disclosed set of price shocks change a browser-local book?

Inputs

  • Local portfolio lotsQuantities, costs, currencies, and acquisition FX retained only in the browser.
  • Current quotesCurrent valuation inputs for symbols present in the example basket.

Calls

curl "https://dhawal.org/api/quotes?symbols=SPY,TLT,GLD"

Transformation

  1. Value every lot with its current quote and the book currency rules shown on Portfolio.
  2. Apply the stated percentage shock to each selected symbol, leaving unselected symbols unchanged.
  3. Revalue the book and subtract the unshocked value.
  4. Keep missing quotes outside both totals rather than treating them as zero.

Output

A before-and-after book value, absolute change, percentage change, and per-symbol contribution.

Limitations

  • The shock is a static scenario, not a probability-weighted forecast.
  • It excludes execution costs, liquidity effects, taxes, and intraday path dependence.
Open productRead methodology
Machine-readable recipe index