Recessions, Indicator-System and Asset-Management Opportunities

In a recent post 1 on econbrowser,  John Kitchen (US Treasury) claims

“based on historical relationships, the recent declines in equity valuations indicate that we are currently in a situation of somewhat heightened risk of a recession – but at this time it is in a gray area that does yet outright indicate a recession compared to what has been observed historically”.

I provided an analysis of the current state of the business-cycle in my previous entry (briefly: the industry sector is subject to contraction, but there is only weak evidence that the sector-dynamics have spread over the whole economy, as of now). Temptingly, in this entry, I’d like to revert the (Granger-) causality implied by John’s statement: instead of inferring recession-probabilities from financial data, I’d like to suggest asset-management opportunities obtained from (macro-) economic indicators.

  • After reader-feedback I here provide the daily indicator series (designs 1-4) for further analysis: indicator_system.txt (the series start in April 1997).

1 Indicator System

The system introduced in my previous entry is made out of four different designs with different intrinsic timing abilities:

  • Design 1 is  lagging (classic growth estimate),
  • Designs 2 and 3 are coincident (either magnify business-cycle or use `faster’ data)
  • Design 4 is a leading indicator (magnify business-cycle and use faster data).

These are daily designs which rely on a novel mixed-frequency extension applied to MDFA. The set of explanatory variables are INDPRO, PAYEMS and UNRATE (monthly), ICSA (weekly) and S&P500 (daily). I emphasize two different target variables: INDPRO-trend and GDP-trend. In comparison to a single-indicator design, a system of indicators with different design priorities offers the advantage of exploiting `typical patterns’ of the system:

  • Pervasive dynamics of all indicators towards protracted downturns.
  • Consistent chronological ordering of `alarm’ signals by the indicators

The next section illustrates results for the INDPRO indicator system.

2. INDPRO Indicator System: an Application to Asset Management

Before starting let me emphasize that:

  • Indicators are based on concurrent (real-time) filters
    • In each time point, filter-outputs rely on current and paste data only
    • In each estimation time point, optimization relies on current and past data only
  • All results rely on the last data vintage. In particular, the history of the monthly macro-data is revised.

Design 1 (lagging)

The following figure shows the history of the first indicator  (red) together with the data (quarterly log-returns INDPRO, black) and the idealized target (symmetric filter, orange:  the latter cannot be used for asset management purposes because it is not a real-time design):

INDPRO

We define trading signals (long/short) according to the sign of the red line and obtain the following trading performances for the S&P500:

perf_design1

Buy-and-hold (green) vs. actively managed (red) and outperformance (blue).

Design 2 (coincident):

Design 2 (blue line below) triggers signals slightly faster than design 1: it is a coincident indicator. Control or manipulation of the timing abilities is obtained by emphasizing that part of the information (of design 1) to which markets tend to be more reactive.

design2

Note that the scale of the blue line is slightly different (which simplifies direct comparisons of all indicators within a single graph…): we can ignore this issue here, since we are interested in the triggering abilities (crossings of the zero-line) exclusively. Trading performances are summarized in the following figure:

perf_design2

The timing of the coincident indicator matches market-dynamics slightly better than the lagging design 1 (annualized sharpe of 0.7 vs. 0.6 for the former).

3. Conclusion

The above results highlighted single-indicator simplistic (test-) strategies which suggest  that there is `meat on the bone’. To dos:

  • exploit the indicator system
  • in the context of more sophisticated asset-management tools (portfolio optimization)

PS: I reduced my exposure to risky assets in the second half of December.

Leave a Reply

Your email address will not be published. Required fields are marked *