Oanda Long Short Ratios

Discussion in 'Forex' started by BenChi, Jun 27, 2006.

  1. BenChi

    BenChi

    Has anyone tried using Oanda's Long-Short position ratios as a sentiment indicator?

    My thoughts are that it should be a great representative sample of what retail traders are thinking, and if we know they are usually wrong, it could be a strong contrarian indicator...

    does anyone have thoughts about/ criticisms of this line of thinking?

    Does anyone have experience putting this to use?

    Thanks for your input!

    -Ben
     
  2. Why not just use commitment of trader figures instead of Oanda's site?
     
  3. ... but the problem is that we can only see the charts. As we cannot download the figures behind the charts, it is impossible to make some reliable statistics.

    As far the COT figures are concerned, I guess that these statistics do not reflect any longer the true nature of the Forex market.
     
  4. BenChi

    BenChi

    Hi Ivanovich,

    To which statistics are you referring? The data seems straightforward enough to me:

    http://fxtrade.oanda.com/resources/trastats/ratio_positions

    The left graph depicts the ratio of long positions to short positions, identifying whether customers are more long or more short, and by how much.

    Yes, COT charts are a valuable resource too, but i'm just wondering what would be the argument against using these numbers?
     
  5. Start a spreadsheet and let us know.


     
  6. The argument against using them is that - unless they represent a significant portion of the trading environment (which they don't), then all they are is "interesting" but not "Actionable".
     
  7. BenChi

    BenChi

    yes, its true that this represents only a small portion of the total open interest, but can we not read it as a representative sample? I guess we would really need some information about the types of people that trade with Oanda to know what the sample is representative of...but unlike COT charts which only represent the positions of the largest few, these numbers should represent a much broader sample of individuals, and in that sense would they not be more statistically valid?

    on a related note, has anyone looked at the 'open position entry price summary information' - the top right and bottom left quadrants represent people holding onto losing trades, and the top left and bottom right quadrants represent people holding onto winning trades....notice which numbers are larger?

    when we combine the information from the position ratio and open position data, we can conclude, for example that of all the Oanda traders trading the USD/CAD, a large number of them are in losing long positions and have sell orders above the current price (perhaps hoping to exit even?). pretty neat!