Read what I just responded to Charles Tines. This strategy is in fact 100% long 5 stocks and 100% short the hedge, even though the developer gives the impression that it’s 50%/50%. I’ve been running this live in a cash IRA account since Jan 1, so I can attest to it. His backtests are done this way, 100% long/100% short with 2:1 margin. I will bet that the long 5 stocks only backtest is also done using 2:1 margin.
1 Like
Excellent response Phil. You are putting your money where your mouth is! Actually walking the walk and now the experience to explain. Thanks!
1 Like
I’ve been thinking about this for a while as well, whether or not what’s being claimed is even mathematically possible.
I think what throws a lot of us for a loop is the reference to a hedge. The word “hedge” makes us think of a cost that must be borne, no matter what. In my world (oil trading), putting on a trade like this would never be called a hedged trade though. It would be called a spread trade. I’ve seen spread trades that produced greater directional moves than either of the underlying legs of the spread, over the same time frame. Something similar could be going on here, depending on a few factors. To test this question, I’ve put together a simple simulation in Excel. I walked through 5 periods, inputting a series of random returns in each period, separately for the long stocks, the short ETF, and then for the combined long-short portfolio.
Exhibit 1
In this example I assume that the stocks have a beta of 1 with the index ETF. Over these 5 periods the portfolio ends up losing money because it suffers decay each time you rebalance (this is the same decay problem that leveraged ETFs have). The cost of a perfect hedge ends up being 101% of the gain on the stocks.
Exhibit 2
Watch what happens when I increase beta to 2. The “cost” of the “hedge” drops to 50% of the gain on the 5 long stocks. The stocks in fiveHedged’s portfolio are generally smaller cap and likely have Betas over 1.
Exhibit 3
Increasing the beta to 3 drops the “cost” of the “hedge” yet again. It’s unlikely the 5 stocks have an average beta this high, but I just want to illustrate my point.
Exhibit 4
Here is where it gets interesting. If fiveHedged’s cash-flow-valuation and institutional-money-flow stock picking system is genuinely good at picking alpha-generating stocks, then it’s entirely possible that these stocks exhibit bifurcated alpha, i.e. they rise more than the general market on up days, but fall less than the general market on down days; no matter what the market is doing, they attract excess capital flows. For this particular series of returns over these 5 periods, if I use an upside beta of 1.4 and a downside beta of 0.6, I get a “cost” to this “hedge” of 15%. This approximates what he shows in the hedged/unhedged backtest comparison, where the hedged backtest’s CAGR is 15% lower than the unhedged’s. So, I think what he shows in the backtest comparison is, at least, possible. However, things get weirder when I show you the next picture…
Exhibit 5
I changed the return in Period 4 from +1.0% to -1.0%. Look at the bizarro result - we make money on both the longs and the shorts. The cost of the “hedge” is negative; it’s actually a profit. This means that if the stock portfolio has larger upside beta and smaller downside beta, and we rebalance to equal dollar amounts after each period, then depending on the specific sequence of returns you can have a weird situation where both the long and short positions make money. If you’re still thinking of the short index ETF as a hedge then this result will throw you for a loop because hedges are costs and costs aren’t profitable.
@QuantitativeModels and @JohnSnow2019 if you want to sense check my spreadsheet, shoot me a PM and I’ll email it to you. Would love to get some feedback on this. That goes for anyone else who wants to check it.
I think the key takeaways here are that 1) in theory, what fiveHedged’s backtests are showing is at least plausible, 2) the real value driver of this system is the stock picking methodology, not so much the hedging. The short index ETF position’s job seems to be to limit large drawdowns during market collapses, 3) most likely the stocks that he is picking tend to have different upside and downside betas, and we could verify this easily enough by doing an analysis on the trade history, 4) overall performance is highly sensitive to the sequence of periodic returns, and 5) weekly rebalancing has an effect as well.
All that being said it’s still entirely possible (maybe even likely?) that in forward live trading the stock-picking won’t look anywhere near as good as in the backtests. Either way, I just wanted to throw this little analysis out there for discussion, since it’s helped me get a better handle on something that’s been confusing all of us for a while on this thread.
Thats absolutely correct @PhilD1. Hedging is only a means to reduce risk. Options are another means to reduce losses. There are numerous methods to hedge.
Its interesting read but I don’t think overanalyzing this is necessary as hedging has been around since derivative products have been around. Institutions use hedging to reduce risk thats why the products were first introduced.
ie. Futures have been around initially to reduce the risk of commodity prices losing value such as soybeans and orange juice…lol
1 Like
Got me thinking about this. Frozen concentrated…
1 Like
Thanks, that is at least mathematically possible!
But then you would be comparing:
A. 100% LONG and 100% short
v.
B. 100% LONG
[which is fine]
Very helpful posts.
But I think the “long 5 stocks only backtest” must be done without margin for the numbers to make sense.
1 Like
I was thinking of that movie as I typed those words as well…lol 
Quick Note:
Swapped HUM stock for ANTM earlier today (intra-week position swap), as both stocks are closely correlated, while ANTM provides a better [deeper] value position. All healthcare stocks currently suffering/prone to the democrat-momentum/“medicare-for-all” rhetoric, which I am keeping a very close eye/analysis upon (more opportunities likely to unfold over coming weeks/months).
Summary: Current positions (five-stocks) therefore are CRCM, SGH, SIGA, VHI, and ANTM, all hedged as always, by long Russell-2000 inverse ETF (RWM).
Five Hedged,
Why did 5 hedged liquid go private?
2 Likes
Mostly practical reasons - I wanted to focus on just one core strategy on C2 which is the same strategy that I personally prefer to trade (and soon to go TOS)…
Some of the earlier liquidity issues with fiveHedged have been addressed (lower turnover, less slippage - please see my earlier post of March 24th) without compromising the core stock-selection methods. This [arguably] negates the need to run two seperate strategies. The focus on one strat also enables much improved time-commitment (and deeper research, data checking etc.) into the individual stocks I select for the fiveHedged portfolio.
Ultimately results matter more than words…
And so far, this objective and ‘focused’ approach has worked, as I expect it to continue to do so into the future.
The biggest plus I have found is that I am not missing any opportunities within all NYSE/Nasdaq stocks (above $250m market-cap) which rank high in my SmartRank algo - details at https://insight.net - while ensuring I am not swapping all 5 stocks every Monday, but rather around 40% turnover (2 buys/sell, 3 holds on avg per week, which significanly reduces slippage). These intricacies/nuances matter, as experienced investors/traders will appreciate.
1 Like
Best month so far, since launch +6.8% in April with a week to go.
Note: Fully hedged long/short strategy…
Magic! Showing current five-stocks of the week…
New All-Time Record High for Portfolio - Up 1.4% Today, Up 23% in 6 months (Dashboard Data). All trades fully hedged.
1 Like
Well done Five Hedged.
Brand New All Time Record High - Strategy Up +24% Since Launch 180 Days Ago.
Fully Hedged by Short Russell 2000 ETF (RWM)…
1 Like
Phil,
Would you explain “the edge… is 0.3”?
Thanks…
Edge = (percent winning trades * average size of winners) - (percent losing trades * average size of losers)
Edge is also know as expectancy. It can be interpreted as the fraction of each unit of risk you take that you earn on average. Edge of 0.3 means for every dollar you risk on a trade on average you earn 30 cents.
A system with 60% winners and 1:1 winners to losers has an expectancy of 0.2. To me this is the threshold for a decent system. Numbers above 0.2 are increasingly good and numbers below are increasingly marginal. This is my personal opinion, not a textbook definition of what’s good.
Interesting side note: most casino games have very small edges, usually 0.05 or less. Nonetheless they make a ton of money because they control the rules (meaning the distribution of outcomes is fixed) and they extract this slim edge many many thousands of times a day. They also limit how much they can lose on any one bet (betting limits). This all adds up to a very profitable business. HFT systems share some similarities with games of chance.
4 Likes
Best Single-Day Performance to Date - New Record High.
Fully-hedged, pure-alpha stock-selection Strategy…
Thanks for the reply…
Interesting observation about casinos - I’ve basically heard this on several occasions, but this was a good context… Also - would a trader/investor accomplish the casino loss controls with proper money management, or is their methodology (conceptually) different?
What other C2 systems do you like?
You can approximate casino risk management by using fixed stops. That’s basically what the casino does, where they set maximum bet sizes, so that no one losing bet (for the house) can wipe out the house. You do the same thing when you risk a set amount per trade and stick to a fixed stop.
The major difference though is that the casino knows exactly what its long run odds of winning are and what the distribution of outcomes looks like, down to infinitesimal precision, because casino games are artificially constructed games with man-made rules. Financial markets are obviously not like that and probabilities and characteristics shift and mutate all the time.
3 Likes
