When knife catching is not knife catching at all, but simply a short-term trading strategy.
We’ve tweaked this strategy as the years have passed and here is quick outline of the main points of sector support buying:
- For the most part we prefer to get involved with ETFs over stocks, or at least put the size on the ETFs and smaller size for the stocks. This type of trade boils down to confidence, and for us it’s much easier being confident in an ETF than a stock.
- This one is very important — it can’t just be one stock hitting support, it has to be the whole sector. The more stocks hit support at the same time, the closer you are to a bounce.
- Best time to add is the day after a trend-day down. The sweet spot for buying support is in the morning sell-off after a trend-day down. That’s when the best bounces come. However sometimes market ends at the lows (this is what happened to miners on Thursday) and then gaps up on Friday. This is the reason we partial in on different days, because it’s very difficult to time the exact bottom.
- Best support buys are when it’s not on fundamental news. For example we wouldn’t buy support on cloud stocks after bad earnings. Period. We stayed away from buying the dip after Japan nuclear crisis because we couldn’t “game” the situation (too bad, bounced great). But buying “margin pukes” is probably the BEST time to buy support as there is no change in the fundamental scenario.
- You don’t anticipate this type of trade, patience is key. You have to wait until there’s blood and then partial in — and never all in on one day. It’s all about buying power and allocation — you have to give yourself concrete levels for adding. As long as the rubber band stretches past support, you can add. Once you’re in overshoot territory your adds will lower your cost basis and eventually your target will be your first entry at initial support. The first target of overshoot is always the first support. For example let’s say you want to get into an ETF called OO. The sector is deeply oversold and many stocks in the sector are hitting support at the same time, similar to what happened here. You start your first partial at 50 which is long term support, and add near the close at 48. You’re now in overshoot territory. Next day it opens at 46, and you add more. Sector starts bouncing. Since you’re in overshoot the primary exit for first partial now is the support that you first started buying, 50, but your cost average is now 48. However, it can also happen in which we are not “feeling” it and do not wait until target and take the profit earlier. Either way, once you get the bounce, your stop automatically becomes the low, no matter what.
- The exit plan if things go badly: if the stocks start basing over the next few days, then you need to start lightening up and taking losses as any basing diminishes the rubber band snap back effect. As said, once the bounce comes that low becomes the stop. If the bounce is not enough for you to get out of your positions profitably and you go back to the lows then you will have to take the loss, which can be quite significant. We hate to jinx ourselves but, to date, after 14 yrs of trading, this has never occurred in which we bail on the entire position without a bounce for significant losses. There is always a bounce in ETFs. However, what can occur (and has happened to us in which we have scratched the trade or gotten out with some damage) is that you get out on the initial bounce but with losses when sector bounces weakly and you feel like it’s your best chance to exit before your position goes back through the lows .
- One last thing: for 90% of our trading we will not go into a trade unless we feel that our risk reward is around 3:1. This means if we risk 1 point we expect to make 3 points. This is very common for traders and is a very sound strategy. This means that even if you’re wrong 50% of the time, you still are profitable due to the 3:1 ratio. However, for this particular strategy this is not the case. Sometimes our profits are less than how much our unrealized losses were but we still consider it a good trade. For example, let’s say we start buying on support and add on overshoot. At the lows we are down 40K unrealized. Sector bounces and we get our first target and start partialling out, bounce stalls and we exit everything for 20K profit. Now most traders will tell you this is not good trading as you were down/risked 2x more than your profits. However, it’s the way it is in the strategy and we haven’t been able to change it — what makes us override this red flag is the extreme consistency of wins the strategy yields. The only time we can endorse a strategy in which the risk is more than the reward (for example 2x) is when the strategy has an incredible win rate. Again, this could be a turn off for many of you, and that’s understandable. This strategy is definitely not for everyone.
There are three places where traders often make errors:
- Getting in too early. They put on a small position early and then see it become quite red and start adding. Don’t even start that original position until there’s blood in the street. Disasters start slowly and this is a prime example of it. They use up all their buying power and cannot lower their cost average. When the bounce occurs it’s not strong enough for them to go green and they exit with losses. This is THE most common error a support trader can make. A two per cent pullback on a strong momentum stock could be a nice entry, but only if you get in on reversal with stop under — we would never use the described strategy for stocks near their highs. It has to be a fast (the faster the better) deep pull-back, often when stocks are hitting the 200SMA or at least 100SMA. Even the 50SMA often is not enough for us to enter this strategy (but yes on individual support buys which are very different than what we’re describing here — for those we wait for reversal, buy, and put stop on that low. We never add and we always have a defined stop). Additional nuance: we don’t always wait for the ETF to hit major support (even though it has to be very close) but the leading stocks in the sector.
- Buying a broken stock instead of buying oversold into support on longer-term bull trend. We would never get into a broken stock just because it’s “cheap”. In our business nothing gets cheaper faster than an already cheap stock. When we say we’re buying on “support”, it automatically means that the long-term bull trend is intact. If it’s a broken chart, by definition, there is no support.
- Not waiting for whole sector to hit support. Probably the main reason we’ve had such consistent success with this strategy is that we simply wait for whole sector to hit support simultaneously. If you want a recent example look at the charts from last week’s post here.
The ETFs that we use most frequently for this strategy are OIH XLE XME GDX GDXJ SIL SMH QQQ SPY, and when it’s with the Ags then POT CF MOS AGU as basket.
We’re primarily break-out traders but there is no strategy that has rewarded us more consistently over the years than buying baskets of deeply oversold ETFs hitting simultaneous support while in longer term bull trends. We hope these detailed notes explain how what some perceive to be “knife catching” is not knife catching at all, but just a good short-term trading opportunity.
HCPG represents a team of three traders who started trading in the late 90s. They started the HCPG newsletter (which shares their next day game plan with members) in 2006. More information can be found at www.highchartpatterns.com and you can follow them at www.twitter.com/hcpg