The publication for trading and investment professionals

eNewsletter

21 June 2012

EURUSD Breaks Long-Term Trendline: What next?

George Davis, chief technical analyst at RBC Capital Markets in Montreal, gives his outlook for euro/dollar as uncertainty around the eurozone continues despite Greece’s election result last Sunday. Davis explains how the medium-term technical outlook reflects declining sentiment in the euro as the currency’s underlying problems persist.

View the online interview >

With the EURUSD lifetime trendline having recently been broken, the way is paved for further declines - an outlook supported by a major chart pattern breakout - which could potentially see the currency retest its all-time low. However, oversold indicators suggest the currency should enjoy a minor fillip in the short-term.

INDICATOR FOCUS:
The Daily Sentiment Index

The Daily Sentiment Indicator (DSI), devised by Jake Bernstein in 1987, is used by institutions on Wall Street as a short-term guide to possible market tops and bottoms. Covering 36 markets including stocks, FX, bonds and commodities, it is published daily and is based on a poll of private traders. As such, it works as a contrarian indicator working on the assumption that private traders are almost always on the wrong side of the market at short and longer-term extremes i.e. tops and bottoms.

Rick Bensignor (pictured), chief market strategist at Merlin Securities in New York, uses the indicator on a daily basis to give him a feel for which markets may be at or near a turn. He stresses however that he never uses the DSI in isolation. “I never use the DSI by itself, nor should anyone else. It merely flags up a market that may be overbought or oversold. I then look at other indicators, such as Tom DeMark's, to confirm a possible turn."

The bullishness is measured as a percentage of “bulls”. Bensignor uses 15% and below as the oversold threshold level and 85% and above as the overbought level. Because it is a contrarian indicator, a low reading is bullish and a high one bearish. Table 1 shows the DSI for June 12-18 (see Figure 1 below).

Looking at the S&P500, Bensignor has noted that the DSI has approached all time lows in recent weeks. The all-time low for the S&P was 3% reached on March 10, 2009. It fell to another low of 5% on October 3, 2011. With its DSI having recently fallen back to 5% again, the S&P has recovered in recent days. “A market needs to remain below 15 or above 85 for several days for the reading to be meaningful” he says. “There may then be a lag of up to a few weeks before actual prices change direction”. (See Figure 2 below).

So, which markets are currently at extreme levels? “Coffee is the one to look at for the moment”, says Bensignor. “It is currently at a low of 8% so may be set for a rebound. The euro recently spent some time below 15 but is now back in the twenties. These are the “raw” readings but don’t forget to look at the moving averages as well for a better idea of the long-term trend”.

Click on thumbnails below for
Figure 1. Daily Sentiment Index of June 12-18
Figure 2. DSI vs S&P500
Table reprinted with permission of www.jakebernstein.com.

A Market Timing Model
for Equity Indices

J.P.Morgan Cazenove’s Global Quant Strategy team headed by Marco Dion in London has developed an interesting market timing model for equity indices that achieves annualised returns of 11.4% in backtests.

The model is designed to take advantage of short-term mean-reversion whilst also taking account of longer trend trends in equity indices. The premises of the model are threefold: i) that markets have been extremely macro driven and directional in nature since the 2007/08 crisis, making stock picking less crucial to successful portfolio management; ii) that markets are mean reverting in the short-term because investors tend to be very emotional and push prices up or down too much; and iii) all relevant information can be found in prices already - “we think that the use of macro-economic information/data is futile when creating a market-timing model”.

The model, which rebalances weekly and which meets specific tolerances regarding maximum drawdown and losing streaks, is based on six indicators, four of which are used in a contrarian manner: The VIX index; Bond Yields; AUD/CHF; and the Dispersion of Valuations in the Market (i.e. how cheap are cheap stocks and how expensive are expensive stocks). The other two signals are market trend (a measure of 3 month momentum) and market seasonality (which is based on their observation that stock markets tend to outperform in the first week of the month).

Each of the six indicators either gives a long or short signal and the combined ‘vote’ then determines the position and position size: +4 or +6 leads to 100% long position; +2 leads to 50% long; 0 is neutral/no position; -2 leads to 50% short; -4 or -6 leads to 100% short. It was important to them to have an even number of indicators such that a neutral “I don’t know” stance could be reflected in the model.

A simple stop-loss is added to the model, which says that as soon as the cumulative equity curve hits losses equalling 2 standard deviations, all open positions should be closed and all subsequent positions avoided. Positions are only re-initiated when the model’s cumulative equity curve returns to its long-term average.

For the MSCI Europe, their benchmark, the model achieved 11.4% annualized returns with a Sharpe ratio of 1.02 and a maximum drawdown over the last 18 years of only -10.4%. They found that their model "still survives the implementation of the trading signals even after transaction costs”.

Please contact Marco Dion to request a full copy of the paper, “Market Timing Model: Methodology, Thinking Process and Strategy to time the market”, (June 2012): marco.x.dion@jpmorgan.com

T: +44 (0)1483 573150 E: enewsletter@technicalanalyst.co.uk

© 2012 Global Markets Media Limited. All rights reserved. Neither this publication nor any part of it may be reproduced, stored in a retrieval system, or transmitted in any form or by an means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of Global Markets Media Limited. While the publisher believes that all information contained in this publication was correct at the time of going to press, they cannot accept liability for any errors or omissions that may appear or loss suffered directly or indirectly by any reader as a result of any advertisement, editorial photographs or other material published in The Technical Analyst eNewsletter. No statement in this publication is to be considered as a recommendation or solicitation to buy or sell securities or to provide investment, tax or legal advice. Readers should be aware that this publication is not intended to replace the need to obtain professional advice in relation to any topic discussed.