Skip to main content
Track Intermarket analysis: bonds, stocks, commodities and FX
1 / 33
Search the archiveLibrary
Library

1990issue C021-5

Policy-auction spread as a weekly equity regime filter

A weekly equity holding can be treated as a stay-exposed or stay-out decision. The archive workflow smooths the policy-auction-spread, converts the sign of the smoothed-spread into a positive-regime or negative-regime state, and checks those states against historical equity paths and open-signal-drawdown.

  • The weekly policy-auction-spread is the administered discount rate minus the three-month bill auction yield.
  • That difference is refreshed as a smoothed-spread with a 15-week exponential average that applies 0.875 to the previous average and 0.125 to the newest difference.
  • The long-side rule starts when the smoothed-spread turns positive and stays in force until that average turns negative.
  • Editorial reading: evaluate the regime-overlay by how cleanly the two states separate historical equity paths and open-signal-drawdown, not as a next-week price forecast.
Entries in this reading3 entries
Current trackIntermarket analysis: bonds, stocks, commodities and FX194 readingsRule-based entry: conditions you can write down and test292 readingsMoving average: simple vs exponential, and what it is for596 readings

A weekly holding as a regime decision

The archive workflow places one equity holding inside a rate-and-policy context. It does not ask the next sampling interval to be predicted from the latest price.

Editorial reading: use the sign of the smoothed-spread as a stay-exposed versus stay-out choice. Judge the procedure by how cleanly those states separate historical equity paths and open-signal-drawdown.

From policy-auction-spread to smoothed-spread

The weekly input is the administered discount rate minus the three-month bill auction yield. That difference is the policy-auction-spread.

A market bill yield below the administered policy rate is treated as the bullish intermarket state. The reverse is treated as the bearish state.

That difference is then smoothed with a 15-week exponential average that applies 0.875 to the previous average and 0.125 to the newest difference. The result is the smoothed-spread.

A worked weekly update used a 7 percent discount rate, a 7.9 percent bill yield, and a prior average of -0.70 to obtain a new average of -0.73. That new average remains below zero, so the update stays in negative-regime.

The long-side rule

The long-side rule starts when the 15-week average turns positive and stays in force until that average turns negative.

Positive-regime is the state in which the smoothed-spread is above zero and treated as the stay-exposed equity window. Negative-regime is the state in which the smoothed-spread is below zero and treated as the abstention or defensive window.

How the historical check was recorded

The historical check used weekly updates from 1953 and recorded the S&P 500 path plus the largest interim close-to-close decline while each long signal was open. That largest interim decline is the open-signal-drawdown.

The two-state wealth comparison assumed an S&P 500-matching holding and left out dividends, commissions, and taxes.

The same crossing rule was described as less dependable for exits than for entries, including a later negative-regime interval during which the equity benchmark later advanced.

Editorial reading: keep more weight on entries and on open-signal-drawdown than on treating every move into negative-regime as a finished exit call.

Buy-signal versus sell-signal S&P 500 wealth from the TB/DR weekly filter

A $10,000 stake left in the S&P 500 only while the 15-week TB/DR average stayed positive grew to $375,593; the same stake held only on sell signals shrank to $3,940. The table records every weekly-rule flip from September 1953 through the 18 August 1989 cutoff, so the two paths show how cleanly the sign of the smoothed policy-auction spread separated historical equity outcomes.
A $10,000 stake left in the S&P 500 only while the 15-week TB/DR average stayed positive grew to $375,593; the same stake held only on sell signals shrank to $3,940. The table records every weekly-rule flip from September 1953 through the 18 August 1989 cutoff, so the two paths show how cleanly the sign of the smoothed policy-auction spread separated historical equity outcomes.S&P 500 · weekly TB/DR 15-week EMA regime · 25 September 1953 to 18 August 1989

Source table ends the last sell path as of 18 August 1989 (asterisk). Dividends, commissions and taxes are omitted, matching the article. Dates use month/day/year as printed.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
1 of 33 in the Intermarket analysis: bonds, stocks, commodities and FX track
19921-5 pp.Next on Intermarket analysis: bonds, stocks, commodities and FXElectric utilities as bond-regime contextUtility stocks have historically tended to turn before the broader equity market at both peaks and troughs, a pattern attributed to their close ties to the bond market.
All readings on this track · 33 readings
  1. 1990Policy-auction spread as a weekly equity regime filter
  2. 1992Electric utilities as bond-regime context
  3. 1992Reading the dollar as a rates-regime check
  4. 1992Evaluating weekly intermarket context for equity regimes
  5. 1993Specifying the stock-bond yield gap as a hold-or-abstain regime
  6. 1996Constructing dual-gate bond-fund entries from gold-silver jumps
  7. 1996Name the regime before the sector breakout
  8. 2002Falling prices flip stock-bond confirmation
  9. 2003Four sleeves on one regime board: gold miners, REITs, bills, and equities
  10. 2003Four currency regimes for the yen, loonie, pound and Australian dollar
  11. 2003Read gold through the dollar regime, the hedge spread, and a stop
  12. 2003When deflation flips the stock-bond map
  13. 2003Commodity subgroup regime boards and dual averages
  14. 2003Reading a liquidity regime when gold, bonds, and stocks rise together
  15. 2003A shared weekly checklist for four country funds
  16. 2004Size-and-style sleeves as a weekly regime map
  17. 2004Country closed-end funds shared one average checklist and four regimes
  18. 2004A 2004 four-pair snapshot of a dollar-bloc FX regime
  19. 2005Country closed-end funds as a weekly regime comparison
  20. 2005Weekly regime maps for production-weighted commodity subgroups
  21. 2005Four technology sleeves on one weekly regime map
  22. 2006The Australian dollar as a commodity regime and timing filter
  23. 2008Dual-listing moving averages as a crowd-regime test
  24. 2008Cross-market regime context for a single trade
  25. 2010Dollar index, cross rates, and commodity context for forex targets
  26. 2010Gold and silver forex session candles as metals-regime context
  27. 2012Yield curve regime and equity timing
  28. 2013Yield curve shapes as stock market regime context
  29. 2013Yield spreads as country-specific equity regime context
  30. 2016Credit spreads as an equity cash regime filter
  31. 2016The summer lull is a context error
  32. 2019Financial sector spreads as regime tells around a global stablecoin
  33. 2019The negative-yield regime as an equity intermarket filter
By fieldIntermarket and spreads264 readingsBy year1990142 readings
All 194 readings tagged Intermarket analysis: bonds, stocks, commodities and FX
Also on Intermarket analysis: bonds, stocks, commodities and FX5 readings
2006The Australian dollar as a commodity regime and timing filter1994Constructing gold-mining rate-of-change tripwires for Treasury bonds1996Constructing dual-gate bond-fund entries from gold-silver jumps1989Evaluating smoothed free reserves and funds-discount spreads1991Regime gates for intermarket and spread trades