Skip to content
Ken Capital

Selected work

Investment research, built to be inspected.

Ken Capital is my independent investment-research portfolio. I turn market questions into documented research — sourcing the data, testing the relationship, building the exhibits, and stating the limitations alongside the conclusion.

Ken Zhang · Baruch College, Zicklin School of Business · BBA in Economics, expected December 2026 · Passed all three CMT Program examinations (charter pending qualifying experience) · CFA Program Level I candidate.

Review selected research View résumé Contact Ken
Dated analysis
Each read carries the date it was made, separate from when it was published.
Source-linked data
The bars and series behind every exhibit ship as downloadable files.
Reproducible exhibits
Charts are generated from the raw data by code, never drawn by hand.
Limitations stated
The caveat — and what would change the read — sits beside the conclusion.

Selected research

All research →
Monthly federal funds rate, 2-year Treasury yield, and 30-year fixed mortgage rate from January 2000 through December 2024.

Featured · Rates · econometrics

The Curve Prices the Fed — the Short End, Not the Long

Question
Does the Treasury curve price the Fed — and where on the curve does the funds rate actually live?
Finding
The funds rate is cointegrated with the 2-year Treasury but only spuriously related to the 30-year mortgage; in the mortgage model, once the 10-year is included, the fed-funds coefficient collapses to 0.005 (t = 0.1).
Method
One regression run at three points on the curve, with cointegration and residual diagnostics — including rejecting an attractive but statistically spurious long-end result.
Caveat
Observational macro data: this is an association the curve prices, not a controlled causal experiment.

Published Jul 2026 · 24 min read

Read full analysis Methodology workbook (.xlsx) Original regression (.xlsx)
SPY daily price making a higher high into September 2024 above its moving averages, over 14- and 30-day RSI panels that each make a lower high — a bearish momentum divergence.

Technical analysis · market internals

SPY, September 2024 — A Higher High on Weaker Momentum

Question
Was September 2024's new high in SPY as strong as the tape made it look?
Finding
Price made a higher high while the 14- and 30-day RSI made lower highs — a bearish momentum divergence beneath a rising market.
Method
Recreated from the original thinkorswim study and reproduced to the penny: Wilder RSI and the 50- and 200-day averages recomputed from daily bars.
Caveat
A divergence flags a fragile advance, not a timing signal — the eventual decline came from an exogenous shock, not the chart.

Published Jul 2026 · analysis September 23, 2024 · 12 min read

Read full analysis SPY daily bars (.csv)
Rivian weekly OHLC bars into mid-2023 forming a converging triangle, with volume drying up locally and then returning — the read flagged before the July production beat.

Technical analysis · volume

Rivian, June 2023 — A Volume Read, Weeks Before the Q2 Beat

Question
Could a volume read flag demand building in Rivian before the fundamentals showed it?
Finding
A converging triangle and volume read as early demand, flagged June 19, 2023 — two weeks before the Q2 production beat that ran the stock to roughly a double off the June low.
Method
A dated weekly OHLC-and-volume read, deterministically aggregated from frozen daily bars; the trade is disclosed as paper-traded, with no entry price.
Caveat
Paper-traded — one worked example of a process, not a track record or a hit rate.

Published Jul 2026 · analysis June 19, 2023 · 15 min read

Read full analysis Rivian daily bars (.csv)
The Shiller CAPE ratio from 1871 to 2026 as a single line, rising to a level near the top of its 155-year range inside a shaded historically-expensive zone.

Valuation · macro

The Shiller CAPE — A Long-Term Map of Where Valuations Sit

Question
Where do U.S. equity valuations sit against their own long history?
Finding
At the July 17, 2026 snapshot, the cyclically-adjusted P/E was the highest level outside the 1999–2000 dot-com era and above the pandemic-era peak.
Method
A 155-year published Shiller CAPE series plotted from the shipped observations, with the mean and prior-peak references recomputed from the CSV.
Caveat
A long-run map of where valuations sit, not a timing signal — expensive can stay expensive for years.

Published Jul 2026 · 4 min read

Read full analysis Shiller CAPE data (.csv)

How the work is built

  1. 01

    Question

    A specific, testable market question.

  2. 02

    Source & document

    Pull the data and record its provenance.

  3. 03

    Test / reproduce

    Run it, and reproduce the original to the number.

  4. 04

    Visualize

    Build the exhibit from the data, not by hand.

  5. 05

    Conclude

    State the finding, the caveat, and what would invalidate it.

The standard each note is held to lives on the process page.

Professional snapshot

  • Economics undergraduate at Baruch College's Zicklin School of Business (BBA in Economics, expected December 2026).
  • Passed all three CMT Program examinations (charter pending qualifying experience); CFA Program Level I candidate.
  • Junior Investment Analyst Intern at Wealth Mill; published markets and economics writing through The Ticker, Baruch's student publication.
  • Based in New York. Ken Capital is an independent investment-research portfolio — personal research, not investment advice.
Read all research View résumé Discuss the research