I presented this idea on another board so I thought I would present it here as well.
My idea is that there is no reason to shift the cloud x periods ahead.
First, let's examine what we are looking at when we shift the cloud x periods ahead.
We are looking at 2 moving averages (highest high + lowest low / 2). When price reaches the cloud, we are observing prices reaction to two moving average prices from x periods ago.
Now, shift the cloud even with the current price. If we want to see price reacting with 2 moving averages from x periods ago, we can look to the Chikou Span and it is giving us the same exact information that we see when the cloud is shifted forward. The only difference is that I am referencing Chikou (prices close) instead of the candles. Does that make sense?
I also believe that using 4 moving averages is a little much but that's just me. The simpler the solution, the better. Less lines, less clutter, but the same information is present. (Occham's Razor)
Like Linkon, I recommend using periods that match the time frame you are viewing. i.e. 22 periods for an Hourly chart (Length of time index/currency futures are open in the US each day)
Please share your thoughts/constructive criticism.
For this example, I am showing 9,26 (17.5 for SSA, and 52 for SSB) settings.