Generalized Technical Analysis. Effects of transaction volume and risk

TL;DR

The paper addresses the limitation of price-only momentum indicators by incorporating transaction volume into a generalized momentum framework. It introduces a volume-weighted generalized momentum indicator that uses $V(t)/\langle V\rangle_{\tau}$ to weight price changes and defines the aggregate form $\widetilde{R}^{\Sigma}_{\tau}(t)$. An investment strategy based on buying at generalized momentum minima and selling at maxima is tested against the classical momentum approach across NASDAQ, NYSE stocks, major indices, and Gold for 1997–2001, with a risk-control parameter $m$. The findings show that the volume-weighted GMI strategy delivers higher returns than the classical strategy under equivalent trading activity, demonstrating the practical impact of including volume information in technical analysis.

Abstract

We generalize the momentum indicator idea taking into account the volume of transactions as a multiplicative factor. We compare returns obtained following strategies based on the classical or the generalized technical analysis, taking into account a sort of risk investor criterion.

Figures (7)

• Figure 1: Daily closing price $y(t)$ for the period Jan. 1, 1997 to Dec. 31, 2001 of stocks traded on NASDAQ -- (a) CSCO, (b) SUNW, (c) AMAT, (d) MSFT, or traded on NYSE -- (e) GE, (f) AOL, (g) PFE, (h) GFI; three major financial indices -- (i) NASDAQ, (j) S&P500, (k) DJIA; the price of Gold (l)
• Figure 2: (a) Volume of transactions $V(t)$ (in 10 Millions) and classical momentum indicator (CMI) $R^{\Sigma}_{\tau}(t)$ for $\tau$=21 days, Eq. (2), for General Electric (GE) for the period Jan. 1, 1997 to Dec. 31, 2001; (b) Reduced volume of transactions $V_r(t)$ and generalized momentum indicator (GMI) $\widetilde{R}^{\Sigma}_{\tau}(t)$ for $\tau$=21 days, Eq. (3), for General Electric (GE) for the period Jan. 1, 1997 to Dec. 31, 2001
• Figure 3: Difference between classical momentum indicator (CMI) $R^{\Sigma}_{\tau}(t)$ and generalized momentum indicator (GMI) $\widetilde{R}^{\Sigma}_{\tau}(t)$ with $\tau$=21 days from Jan. 1, 1997 to Dec. 31, 2001 for stocks traded on NASDAQ -- (a) CSCO, (b) SUNW, (c) AMAT, (d) MSFT; stocks traded on NYSE -- (e) GE, (f) AOL, (g) PFE, (h) GFI; three financial indices -- (i) NASDAQ, (j) S&P500, (k) DJIA, and Gold (l)
• Figure 4: Comparison of returns (Eq.(\ref{['return']})) in per cent between the generalized momentum indicator GMI-strategy and classical momentum indicator CMI-strategy for different values of $m$-criterion for (a) Gold and (c) GE. Comparison of the cumulative number of recommended transactions $N_c$ performed when GMI-strategy and CMI-strategy are implemented for different values of $m$-criterion for (a) Gold and (c) GE
• Figure 5: Retuns obtained following (GMI)-strategy (full diamonds and solid line) and CMI-strategy (open diamonds and dotted line) for $\tau$=21 days and restriction policy parameter $m=$0.01 for the period Jan. 1, 1997 to Dec. 31, 2001 for stocks traded on NASDAQ -- (a) CSCO, (b) SUNW, (c) AMAT, (d) MSFT