Educational use only. SAVNG is a research tool, not investment advice. Models can be wrong. Past performance does not predict future results. Verify every figure before risking capital.
Don't outsource your judgment

We tested 69,000 insider purchases
and value 3,500 companies from their own filings.

Twenty years of SEC filings, eleven market regimes, every company counted once. Here is what holds up, what does not, and what we got wrong — with the stocks each finding points to today. Look up any ticker, or start with the research.

Tested screens, today's market, and the stocks they point to

Every screen on this site carries its own twenty-year record and the list of stocks that meet it now. Start with the ones that survived an out-of-sample test.

Screens that survived an out-of-sample test

We searched 518,320 rule combinations on filings before 2017, then checked the survivors on filings from 2017 onward that the search never saw. 11 held up; 39 did not.

P/B ≤ 0.6 · non-financials+41.2 pts
P/B ≤ 0.8 · non-financials+28.5 pts
P/B ≤ 0.8 · under $2B · non-financials+35.8 pts
P/B ≤ 0.8 · F-score ≥ 5 · non-financials+31.1 pts
12-month return vs the S&P 500 in the confirmation years. Each screen opens with its full record and the stocks that match today.
Open the Screen Lab →
Where the market is today
Calm
S&P 500 3.1% below its 1-year high
0 / 6
top-warning signals in their danger zone

Sell-off = 10% or more below the 1-year high, the same rule every backtest here uses. The monitor scores every warning signal we hold on twenty years of tops, including how far the market rose after each warning.

Market-top monitor →
How this site works
  • Every claim has a record. Screens, insider signals and the equity-bond lens are tested point-in-time on SEC filings back to 2006, with the misses published.
  • Every screen has today's list. Run any rule and see the stocks that meet it now, next to how that rule did before.
  • Every stock has a valuation from its own filings. 3,500 companies, one method, no analyst opinions.

What twenty years of filings actually show

Tested in public, each company counted once, graded against the right benchmark. Misses and corrections published next to the hits.

Market risk — right now
Across the 3,538 stocks we can value right now
⚡ Right now ?Near-term risk over roughly the next 3 months. Blends live market conditions (volatility, how far the S&P is from its highs, credit spreads, the yield curve, sentiment), how stretched valuations are, and what the smart money is doing. Higher = greater odds of a pullback. · next 0–3 months
44 /100 Elevated ▼ -5 mild
Higher than 58% of the last 67 days we've tracked.
Long-term risk ?Long-term / structural risk — the market's underlying tide. Where valuations sit versus 145 years of history (Shiller CAPE) plus multi-year trends. A high reading means weak long-run return prospects even if the market feels calm right now. · structural · vs 145yr history
65 /100 High ▲ +1 slight
Shiller CAPE 40.5× — U.S. valuations are richer than ~99% of the past ~145 years.
▲ / ▼ = change vs ~7 days ago (2026-09-10) — red = risk rising, green = falling.
Inside the "right now" score — a 45 / 30 / 25 blend · 3,538 stocks valued
Conditions · 45% ?The live market weather: VIX (expected volatility), how far the S&P 500 is below its 52-week high, high-yield credit spreads, the 10-year vs 2-year yield curve, and the Fear & Greed index. Market-wide indices, not per-stock.
35 Moderate
▼ -7
Valuation · 30% ?How expensive stocks are versus our intrinsic-value models: the median price-to-value ratio and the share of stocks trading above what we think they are worth, across 3,538 stocks.
52 Elevated
▲ +1
Smart money · 25% ?What informed money is doing: short-selling activity plus corporate insiders buying or selling their own stock. Broken out below.
51 Elevated
▼ -7
Smart money — the three inputs
Short-selling ?Share of ALL US trading volume that is short sales (FINRA, ~12,000 symbols, market-wide). It normally sits near 45–50%, so readings above ~50% lean bearish and read as higher risk. Weighted 40% of Smart money.
52.0%
Elevated
◆ flat pt
Buyers' strike ?Share of the sampled companies whose insiders bought NONE of their own stock on the open market in the last 12 months. The higher this runs, the more bearish (nobody's buying the dip in their own shares). Weighted 35%.
54%
Elevated
◆ flat pt
Net insider selling ?Share of the sampled companies whose insiders were net DISCRETIONARY sellers over 12 months (excludes scheduled 10b5-1 plans, which carry little signal). Around half is normal; well above that is bearish. Weighted 25%.
35%
Moderate
▼ -6 pt
Black number = the raw figure; coloured bar & word = how it reads as risk. Insider inputs from 4,610 stocks (full universe); short-selling is market-wide.
Driving "right now": short-selling elevated (52% of volume).
Long-term risk: the long-term read runs high mainly because valuations sit near a historic extreme (Shiller CAPE) — a structural backdrop, not about this week.
Scale (0–100): 0–20 Low · 20–40 Moderate · 40–60 Elevated · 60–80 High · 80–100 Extreme. "Right now" reads today's conditions; the "long-term" score is anchored to ~145 years of valuation history.
Don't outsource your judgment

We backtest our own insider signals in public — winners and losers, nothing deleted.

Most "insider buying beats the market" claims fall apart once you stop double-counting and compare against the right benchmark. Ours mostly does too — so instead of selling you a number, we show you the honest scorecard: every signal type graded against the S&P 500 and small-caps, and a public ledger of every high-conviction call, timestamped before the outcome exists.

281
signals logged & graded as they mature
Broad insider buying: +2.9% vs S&P · -2.2% vs small-caps
Beats the market 43% of the time — not a proven edge. The honest scorecard shows where it is and isn't.
Hypothetical event-study returns from public SEC filings and our own models — no trading costs; past performance does not predict future results. Educational, not investment advice. Methodology & disclosures.

Try the valuation calculator

Pick any stock, then change the assumptions yourself. We show what the market is pricing in — and what the business is worth if you're right.

Apple Inc.
AAPL · Technology
Deeply overvalued
Price$312.51
Our IV$115.28
P / IV2.71
Market is pricing in
What if you assume…
Your intrinsic value
Drag a slider to see how the value moves.
Open the full analysis →

This week's economic data, in plain English

FRED, as of 2026-09-17 · The full backdrop →
New jobless claims, latest week
206k▼ from 207k
week of 2026-09-05
206k people filed for unemployment for the first time last week (207k the week before). Layoffs remain low.
Core inflation (CPI without food and energy)
2.8%▼ from 2.8%
August 2026
Prices outside food and energy are rising 2.8% a year, down from 2.8%. Above the 2% goal but not by much.
Inflation (CPI)
3.7%▲ from 3.5%
August 2026
The overall cost of living is rising 3.7% a year, up from 3.5%. Food and energy swing this number; the core reading is the steadier one.
Unemployment rate
4.1%▬ from 4.1%
August 2026
4.1% of people looking for work cannot find it, unchanged from 4.1%. Low and steady is the healthy reading.
Jobs added in the month
162k▲ from 21k
August 2026
Employers added 162k jobs last month (21k before). Solid hiring.
Retail sales, month over month
1.2%▲ from -0.5%
August 2026
Shoppers spent 1.2% more than the month before. Consumers are two-thirds of the economy; a run of negative months matters, one does not.
Yield curve (10-year minus 3-month)
0.87 pts▼ from 0.89 pts
as of 2026-09-16
Ten-year yields are above the three-month rate by 0.87 points. A normal, upward-sloping curve.
Fed funds rate
3.63%▬ from 3.63%
as of 2026-09-15
The Fed's policy rate is 3.63%. Higher rates slow borrowing and weigh on stock valuations; cuts do the opposite.
The underlying readings Regime: Neutral / mid-cycle as of 2026-09-17

No extreme readings in either direction. Stock selection matters more than macro positioning right now.

Shiller CAPE ?Shiller CAPE
Shiller CAPE — the S&P 500 price divided by its 10-year average inflation-adjusted earnings. The longest valuation record we have (since 1871). Readings above ~38 have happened only twice: the 1999 dot-com peak and today. High CAPE historically means weak long-run returns.
Source: multpl.com (Shiller CAPE)
40.5x
Extreme (near record)
Yield curve (10y–2y) ?Yield curve (10y–2y)
10-year minus 2-year Treasury yield, in percentage points. An inverted curve (long rates below short) has preceded every US recession since the 1970s. The 10y–3m spread is +87bp.
Source: FRED (T10Y2Y / T10Y3M)
+27bp
Flat
Credit spread (HY) ?Credit spread (HY)
High-yield credit spread — the extra yield junk bonds pay over Treasuries. Widening spreads mean investors are demanding more for default risk; it's an early stress signal that usually leads equity selloffs.
Source: FRED (BAMLH0A0HYM2)
2.76%
Tight (complacent)
VIX ?VIX
Market expectation of S&P 500 volatility over the next 30 days. High = nervous; low = complacent.
Source: Stooq (^VIX)
15.7
Calm
Fear & Greed ?Fear & Greed
Composite of put/call ratios, breadth, momentum, and safe-haven demand. Contrarian indicator — extreme fear historically marks bottoms; extreme greed marks tops.
Source: CNN Fear & Greed Index
29
Fear
10Y Yield ?10Y Yield
Yield on the 10-year US Treasury. Higher rates compress equity valuations and raise the bar for risk-on bets.
Source: FRED (DGS10)
5%
Tight (high rates)
Inflation (10y exp.) ?Inflation (10y exp.)
10-year breakeven inflation (nominal Treasury minus inflation-protected TIPS yield) — the market's expected average annual inflation over the next decade. It's the silent tax on every nominal return.
Source: FRED (T10YIE)
2.33%
Moderate
DXY ?DXY
US dollar strength vs basket of major currencies. Strong dollar pressures multinational earnings and emerging markets.
Source: Stooq (^DXY)
100.11
Stable
Gold ?Gold
Spot gold. Rising gold often reflects inflation fear or risk-off positioning.
Source: FRED (London PM fix)
$4,406
Per troy oz
Oil (WTI) ?Oil (WTI)
WTI crude oil $/barrel. High oil raises input costs across the economy and pushes inflation up; low oil can signal weakening demand.
Source: FRED (DCOILWTICO)
$107.02
High — inflation risk
BTC ?BTC
Bitcoin as a risk-on liquidity indicator. Rallies often coincide with broad risk-asset rallies; sharp drops can precede equity weakness.
Source: CoinGecko
$76,377
$76,377
M1 supply ?M1 supply
M1 money supply (cash + checking deposits). Falling M1 signals tightening liquidity; rising suggests stimulus is loose.
Source: FRED (M1SL)
$19.9T
$19.9T
Sources: FRED (rates, curve, credit) · multpl (Shiller CAPE) · CNN Fear & Greed · CoinGecko · Stooq. The verdict block on each stock page interprets these in context.

What you get on every stock page

Not a black-box buy/sell signal. A valuation you can rebuild yourself, what today's price already assumes, where the model is weak, and four distress checks — because insider buying does not predict survival.

DCF intrinsic value

Five-model engine (FCF DCF, Residual Income, DDM, Normalized FCF, Growth DCF) auto-routes by sector + data quality. Conservative / base / optimistic scenarios weight into one IV range with margin-of-safety vs current price.

Browse the universe →

NCAV bargain screener

Graham's net-net screen surfaces companies trading below liquidation value. Almost always small caps in distress — risky as individual picks, statistically powerful as a basket. We flag, you decide.

See the screener →

AI news + sentiment

Gemini-summarized recent news, analyst commentary, social sentiment, and three relevant YouTube videos per stock. Lightweight context layer on top of the hard numbers — not a recommendation.

Browse analyses →

Ready to look up a stock?

Every page shows our DCF, anchor comparisons, and AI summary side-by-side. Verify every number yourself before risking capital.

Important. SAVNG is published by Pouyan Golshani, MD — founder of Gighz. All content is for general informational and educational purposes only. It is not investment, tax, or legal advice and does not constitute a recommendation to buy or sell any security. Models can be wrong, data sources can be wrong, and the future can surprise everyone. Consult a licensed financial advisor before risking capital. See full disclaimer.