AI UNDERDOGS
Back
0:00 / 1:00
#008

Contrarian Signals

contrariansignals.com/
FinanceSentiment analysisContrarian investingIndie project
131 views💬 0 comments🔗 0 visits

A Bloomberg-style terminal for contrarian sentiment signals — and it's free

WHAT IT SOLVES

Retail investors trade on emotion, but nobody puts the major sentiment indicators side by side and tells you when to go against the crowd

WHY IT'S INTERESTING

Product taste

Three indicators that actually matter

CNN Fear & Greed, AAII Investor Survey, Bank of America Sell Side Indicator — these three cross-validate retail vs. institutional sentiment. Choosing the right indicators is itself an analytical call

Real craft

8 AM email, no terminal required

Aggregates multiple indicators into a single BUY/SELL/HOLD call, then explains why in plain English. The 5/28 signal reads: 'Retail looks notably pessimistic while broader risk appetite is mildly optimistic. Buy gradually into weakness.' Not just numbers — actual reasoning

TECH GUESS

Bloomberg-terminal-styled frontend, scheduled backend scraping three public data sources, likely a transactional email service for daily delivery — HN title says it was built with Claude Code

DEEP DIVE

Three Curated Indicators, Not a Random API Grab

Contrarian Signals aggregates three sentiment indicators that actually matter in institutional circles: the CNN Fear & Greed Index, the AAII Investor Sentiment Survey, and the Bank of America Sell Side Indicator. This is a deliberate selection, not a "plug in whatever free API I can find" approach. Each one captures a different layer of market psychology — CNN's composite for overall mood, AAII for retail investor sentiment, and BofA's SSI for how institutional strategists are positioned. Triangulating across all three gives you something no single indicator can: a cross-validated read on whether the crowd is wrong.

Natural Language Buy/Sell/Hold Signals — Not Just Raw Numbers

The smartest design choice here is the synthesis layer. Every morning at 8 AM ET, subscribers get an email with a clear BUY, SELL, or HOLD recommendation, plus a plain-English explanation. The May 28 signal read BUY, with reasoning like: "retail looks notably pessimistic while broader risk appetite is only mildly optimistic and institutional positioning sits in the middle. Buy gradually into weakness." This is essentially translating Warren Buffett's famous dictum — "Be fearful when others are greedy and greedy only when others are fearful" — into a daily, actionable digest. No dashboard scrolling required.

70% Vibe Coded, Aesthetics-First Personal Project

In the Hacker News thread (titled "Bloomberg-inspired market sentiment tracker built with Claude Code" by victordg — 3 points, 1 comments), the developer was refreshingly honest: roughly 70% of the project was "vibe coded." The motivation was aesthetic first — he wanted to build something with the dense, information-packed feel of a Bloomberg terminal, and a sentiment dashboard was the perfect excuse. This is a passion project without a business model, which is both its charm and its limitation. The AI coding tool (Claude Code) clearly helped ship the UI and data plumbing fast, but choosing which indicators to aggregate and how to weight them in a composite judgment requires domain intuition that LLMs can't substitute for.

Honest Limitations Worth Knowing

First, all three indicators are US-market-centric — if you trade A-shares or Hang Seng, the signals may not translate well. Second, sentiment is only one dimension of investing; this tool ignores valuation, fundamentals, and technicals entirely, so it should never be your sole decision input. Third, the weighting logic behind the BUY/SELL/HOLD synthesis is opaque — how much does it trust CNN Fear & Greed versus the AAII survey? Without transparency on that, you're taking the recommendation on faith. Fourth, with only 3 points and 1 comment on HN, the tool hasn't yet attracted enough scrutiny or user feedback to validate its signal quality over time. For contrarian-minded developers and PMs, it's an interesting specimen of AI-assisted solo building, but temper expectations accordingly.

📍 Source: hn📅 2026-05-22Original post →Visit site →
Ad
Ad slot (AdSense unit renders here once connected)

Discussion (0)

Sign in with GitHub to post
  • No comments yet — be the first.

Related