SignalStats: Compute Trading Signals Statistics

In JustinMShea/ExpectedReturns: Reproduction of Investment Strategies presented in Antti Ilmanen's Expected Returns and Investing Amid Low Expected Returns

Compute Trading Signals Statistics

Description

Trading signals can be further analyzed with several statistics, across instruments. Signal statistics currently implemented include:

• Activity matrix (method="activity"). Represents the number of long or short positions over the periods considered. It is expressed in percentage units.

• Agreement matrix (method="agreement"). It denotes the number of periods during which signals pairs agree on the position, on average across instruments. It is expressed in percentage units.

• Correlation matrix (method="correlation"). The average correlation among signals, across instruments.

Usage

SignalStats(X, symbols, signals, method, ...)

Arguments

X	A list of xts objects storing signals series. See 'Details'.
symbols	A character vector providing the names of symbols signals were computed for.
signals	A character vector specifying the names of signals provided in X.
method	A string indicating which statistics to compute among X signals. One of "activity", "agreement", or "correlation".
...	Any other pass through parameter.

Value

Varies depending on the chosen method.

Note

Some signals may result into an inactive position. Such positions are not considered in the activity matrix and thus values do not sum up to unity. Whereas, they are accounted for in the agreement matrix.

Author(s)

Vito Lestingi

References

Baltas, A. N. and Kosowski, R. (2012). Improving time-series momentum strategies: The role of trading signals and volatility estimators. EDHEC-Risk Institute.

See Also

MomSignal()

JustinMShea/ExpectedReturns documentation built on Sept. 9, 2023, 9:41 p.m.