[CAMWEST-discuss] The ability to forecast

Danny Hannan danny_hannan at yahoo.com
Mon Feb 9 01:00:13 UTC 2009


G'day all,
This e-mail is in response to a reply to my recent letter to politicians regarding the most recent federal stimulus package; "one rant too many".  I realise that that one reply might not be the only person who feels that way.

I would like to put some background to my articles.  The ability to produce accurate future projections relies on the accuracy of the necessary assumptions used.

There has been a considerable amount of research (multiple studies) into forecasting methods across a wide range of areas of study from science based subjects to fortune telling.  The accuracy of forecasting is dismally poor with one major exception; up to a 14 day weather forecast.  The measure used was:  The period of forecast was compared to the same time period in the past.  The trend or pattern of the past time period was used as a datum line and the forecast was measured against that datum line.  The only forecasting system that was considerably more accurate (statistically) than a continuation of that datum line was weather forecasting up to 14 days.

That means the most accurate assumption to use for a future projection is a continuation of past trends.  Unfortunately many of our professional, financial and scientific forecasting agencies ABARE, IEA, CSIRO, EIA, IMF have political interference with assumptions for projections thrust onto them (the assumptions are too wild to come from experts in the field).  Though the assumptions are published with the report; unless you are expert in the particular field or do the research to confirm or discredit the assumptions the projections are taken on face value and often used erroneously for future planning by both governments and industry.

I do the research for both my own interest and for my income as a private international oil/energy futures trader.  I write the articles to help me clarify my own thinking and organise my thoughts.  I only use past trends extrapolated into the future for my projections.  But I think my articles and letters are useful to all in the community.  If you would rather me just give a web address for the letter or article instead of attaching or putting it in the body of the e-mail please say so.  Doing so requires more work for me so unless I feel there are many who consider my e-mails "rants" I will continue as in the past.

Dan

danny_hannan at yahoo.com


      Make Yahoo!7 your homepage and win a trip to the Quiksilver Pro. Find out more



More information about the CAMWEST-discuss mailing list