© Andrew Coles       Terms & Conditions | Privacy

Modified VWAP Methodologies
Share on Facebook
Share on Twitter
Share via e-mail
Share on Delicious
Share on Digg
Share on Google Bookmarks
Share on LiveJournal
Share on Newsvine
Share on Reddit
Share on Stumble Upon

Essay two continued

Further comments on the five generations of MIDAS curve continued

Second generation curves - a more detailed overview

As stated earlier, second generation curves were vital for two reasons. First, their use is critical in market contexts where volume is either fluctuating excessively or where there are very large pronounced volume trends. In such cases, first generation curves (inputting market volume) can produce unreliable areas of support and resistance. Paul Levine was unaware of these issues. Second, some markets – most obviously the cash foreign exchange markets – do not supply volume and so the entire MIDAS system could not be applied to them. It is true, as I (Andrew Coles) also demonstrated in Chapter 10, that intraday cash forex traders can feed tick data into S/R and TB-F curves. Many traders question the reliability of tick data but the finding of this chapter was that tick data are an adequate intraday substitute for volume. However, for higher timeframe cash forex participants tick data are not available. A related rationale for second generation curves is that market data supplied by some sources don’t contain volume or else they contain unreliable volume data. Second generation curves are again indispensable in such circumstances.

Let’s expand a little on these issues while looking briefly at the following:

  1. Cash forex markets (tick data and higher timeframe volumeless charts)
  2. Forex futures markets
  3. New futures contracts and expiring futures contracts (rollover)
  4. Very long-term secular charts
  5. Data sources without volume or with unreliable volume
  6. Futures open interest

1. Cash forex markets

I (Andrew Coles) discussed this application thoroughly in my Chapter 10. There were two main points in this chapter that cash forex traders should be aware of.

First, despite the concerns many intraday cash forex traders raise about tick data (see in particular pp270-271), I (AC) demonstrated that replacing volume by tick data does not significantly affect the functioning either of M-S/R curves or the TB-F. Indeed, a comparison of both curves in the cash and equivalent futures markets (using futures volume in the latter case) revealed virtually identical curve profiles. Accordingly, tick MIDAS curves are legitimate second generation curves that a cash forex day trader should absolutely consider using.

The second point was that for higher timeframe volumeless cash forex charts nominal/constant volume curves are highly dependable and indeed essential. Chapter 10 is replete with illustrations of second generation curves functioning both as M-S/R curves and TB-F curves consuming nominal volume. By the same token, indicators such as my MIDAS Displacement Channel, MIDAS Standard Deviation Bands, and Bob English’s Detrended Curves are also fully functional in the forex markets provided they’re second generation indicators utilising tick data or nominal volume.

2. Forex futures markets

I (Andrew Coles) included the forex futures markets in this section because these markets exemplify extremely well the problem identified earlier in the discussion of rapidly fluctuating volume. Indeed, the same problem afflicts the intraday cash forex markets with regard to tick data.

As I discussed in Chapter 10 (see in particular my section A Comparison of the MIDAS S/R Curves Using Cash FX Intraday Tick Data and Intraday Futures Volume Data, pp270f), fluctuating intraday tick data and futures volume predictably occur during trading centre overlaps between Asia and Europe and Europe and the United States. In such circumstances, price levels can frequently penetrate MIDAS curves (porosity) or, conversely, terminate just below them (suspension).

While Paul Levine didn’t recognise the problem of price suspension – the converse of the porosity problem – he thought porosity occurred because MIDAS curves were only approximations “to a more complex and less deterministic reality” (Lecture 5). However, in Chapter 13 I (Andrew Coles) discussed several solutions to these problems by David Hawkins and myself. My solution #7 (p341) is relevant here, since, as I expanded this solution fully my Chapter 11, it is vital (a) that the MIDAS analyst understand how volume trends impact the curves, and (b) that the analyst apply the rules in Chapter 11 so that a decision can be made concerning which MIDAS curve – first or second generation – to apply. Indeed, after a time the four rules become so ingrained that an analyst can visualise where each type of curve is likely to fall even without plotting both.

I suspect that many MIDAS users still don’t appreciate the impact that volume trends have on the chart position of MIDAS curves, and still less do they appreciate the need for second generation nominal curves.

3. New futures contracts and expiring futures contracts (rollover)

The drop in liquidity in expiring futures contracts and the relative lack of liquidity in new contracts is well-known during the rollover period. In my Chapter 11 I (Andrew Coles) created four rules for trading price and volume relationships. Two of those rules apply directly to the rollover period when there is lower market volume, namely Rule #2 for rising price trends and Rule #3 for falling price trends.

Applying Rule #2, MIDAS curves tend to be pulled down below the rising price trend, while applying Rule #3 MIDAS curves tend to be pulled up above the declining price trend. In other words, there is price porosity in rising price trends and price suspension in falling price trends.

In such circumstances, it’s again crucial to use second generation nominal curves, since on the whole they tend here to produce much more accurate support and resistance than first generation curves. In due course I will illustrate this in a blog post.

previous page next page