6a0105370026df970c019102d5eb31970c-800wi
Source: Stockcharts

 

 

Definition: Hindenburg Omen is triggered when: (1) more than 2.2% of stocks on the NYSE are at 52-week highs AND more than 2.2% are at 52-week lows, (2) the 50-day moving average is trending higher, (3) the McClellan Oscillator is negative, and (4) new 52-week highs don’t exceed new lows by more than 100%.

 

Yesterday, we asked the question Why people fear the Hindenburg Omen? despite its mixed (i.e., awful) track record in forecasting crashes.

Today, I have a more specific technical question: With markets up 16% YTD as of mid-May, we obviously had a huge number of stocks making 52 week highs.

But we also have seen a spike in rates, suggesting that anything credit or bond related — closed end funds, mortgage REITs, bond funds, etc. — trading on the NYSE are going to be trading appreciably lower. Perhaps even to 52 week lows.

Then there is Japan, and the large number of ADRs and ETFs on the NYSE. These may also be trading lower; the same for various European ADRs that are faltering.

My technical question is this: What happens if we look at the NYSE US operating company only for this indicator — does it change its track record? Does removing the non equity names improve the signal’s otherwise mediocre track record?

Alternatively, when successful Hindenburg Omens have been made in the past, are bonds/ADRs/closed end funds/non operating companies a factor?

Is it possible to generate a track record using Nasdaq as the basis instead of the NYSE? What about the 500 names in the S&P500.

Inquiring minds want to know!

 

Hindenburg-disaster

Category: Markets, Sentiment, Technical Analysis

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

5 Responses to “Question: What Is Driving Hindenburg Omen Internals?”

  1. Disinfectant says:

    I was pointing out this issue on several financial sites back in 2010. I believe that most of the supposed Omen sightings over the past decade are in error due to reliance on NYSE data instead of relying on a pure stock dataset. Since stocks and bonds have been generally negatively correlated, we have had various instances where one of the two are hitting highs and the other is hitting lows, setting off the Omen. But that is not the logic as to why the Omen is supposed to work, which is based on divergences within the stock market, not divergences between stocks and bonds (which is routine).

    2008, however, was likely a legitimate “hit” since there was in fact a divergence in the stock market in the first half of the year; most stocks were falling, but commodity-oriented stocks were shooting higher. Thus, it is actually possible that the Omen is a much better indicator than assumed, but one would have to use stock-only data.

    Barry, you should check with Ned Davis on this as they track stock-only indexes and they may have already analyzed this matter. If not, they should be able to do it fairly easily.

    • Disinfectant says:

      P.S. Based on stock-only data, I don’t think a legitimate Omen has been triggered this year, but I’m not 100% sure of that.

  2. Excellent questions. I for one will be grateful for folks taking a closer look, and I doubt I’m the only one! 3 comments…

    (1) The lack of published articles about the Hindenburg-equivalent in other markets suggests that the NYSE may be unique in structure for some reason. Your questions about Japan and so on are excellent but unless someone has already answered them, getting answers might take time… The metrics in the Hindenburg signal (the McClellan Oscillator and New Highs/New Lows), rely heavily on number of issues, and signals will be sensitive to which types of issues are in the market.

    I would comment that the NYSE has had closed-end bond funds for decades. The current market structure using ETFs is different than the historical structure, but perhaps not so different… The current NYSE listings page, when filtered for Closed-End Funds, returns about 400-500 issues, of which the majority are bond funds.

    http://www.nyse.com/about/listed/lc_ny_issuetype_1076458359980.html?ListedComp=All&start=21&startlist=1&item=2&firsttime=done

    For historical data, I happened to find an InsideNYSE newsletter which mentioned “As of June 2007, the NYSE trades 497 closed- end funds with a market cap of approximately $220 billion.”

    http://www.nyse.com/pdfs/InsideNYSE2Q07.pdf

    (2) Perhaps, in addition to looking at what is on the new-lows list during Hindenburg signals, one should also look at what is on the new-highs lists?

    (3) John Hussman’s work shows that strong market declines occur following periods when there are rising interest rates (bond yield metrics) in the context of an overbought (e.g. price hitting bollinger bands) and overbullish (II advisor survey data showing high % bullish) market. So it would not be surprising if rising interest rates are part of why the Hindenburg signal is present before major market declines.

  3. Here is link to NYSE New Highs / New Lows historical data via WSJ. You can change the date string in the URL to target a given date. The link below is for yesterday’s Hindenburg signal (June 4, 2013; 88 new highs and 134 new lows). This online dataset goes back to at least May 1, 2007.

    http://online.wsj.com/mdc/public/page/2_3021-newhinyse-newhighs-20130604.html?mod=mdc_pastcalendar

    Scanning the new lows for June 4, I see Muni bond funds, some other bond funds, REITs, and preferreds. Rate-sensitive issues. A few ‘net stocks (Red Hat, Rackspace Hosting). Very few ADSs. No Japan anything to my eye.

    New Highs are a wider mix, heavily stocks. Ford, GM and its warrants; Target, Tootsie Roll, Skechers, the LyondellBasell that was just added to ValueLine, CIT… nothing jumps out.

    Now looking back to May 31 (the confirmation of the May 29 sighting) when the New Highs were 98 and New Lows numbered 162. New Highs are a mix of stocks again, lots of industrials. Some preferreds too. New lows are dominated by BlackRock, Nuveen, and PIMCO muni bond funds. But again, no Japan.

    I think we can rule out Japan, though I still think the question of whether Hindenburg-type signals are meaningful in other indices is worth looking into, if anyone has the data.

    For historical comparisons, I went back to the October 16, 2007 Hindenburg, around the time of the last major market peak. New Highs (75) were dominated by stocks, similar to the analysis above. New Lows (81), IN CONTRAST TO TODAY, were dominated by stocks as well. Familiar names on the 2007 new lows included Walgreens, Toyota ADS, Talbots, RubyTuesday, Regions Financial, JCP, McClatchy, LaZBoy, Nautilus, NY Times, Nordstroms, Briggs & Stratton, Dominos Pizza… There were a healthy number of mortgage lenders and financials, plus some of the early Nuveen Muni ETFs listed too, but not as many as we’re seeing in 2013.

    I also checked the December 14, 2010 data. That was during a time when munis were suffering (remember Merideth Whitney going on 60 minutes warning about municipal default risks being much higher than the market appreciated). So that was a time when the Muni issues on NYSE were hitting new lows, but for that Omen, the market as a whole did okay (perhaps all the QEasings helped)…

    So it looks like the question of whether the Omen’s accuracy can be improved by focusing on US stocks vs. other issues is a very good one. Particularly because stocks are now well along from the bear market low, anticipating the next major top is a major concern.

    A full listing of historical Hindenburg occurrences back to 1986 is given by Robert McHugh here. He uses a 2.2% threshold for the New Highs and New Lows, which is lower than the 2.8% used by Miekka… but requires a confirming second instance before counting the Omen off.

    http://www.safehaven.com/article/30019/we-got-an-official-confirmed-hindenburg-omen-on-may-31st-2013

  4. Patrick Neid says:

    Sustainable Gains, thanks for the link at Safehaven. This from the article:

    ……”Except for that one crash, no stock market crash (a decline greater than 15 percent) occurred over the past 29 years without the presence of a Hindenburg Omen. Another way of looking at it is, without an official confirmed Hindenburg Omen, we are pretty safe. On the other hand, if we have an official Hindenburg Omen, then a critical set of market conditions necessary for a stock market crash exists. As of May 31st, 2013, we have such a condition in the market, as we have a new official Hindenburg Omen.”

    This why I think the Omen, like the grim reaper is not a good sign! The chess game continues.