Request for a starting point on Sector Rotation Strategies

Hi,
I had been a fairly active participant in the Wealthlab community till about 3 years ago, when life decided to pull me in a different direction. Then last week I try to login there and the website and the looks are completely different. So, I was looking for a similar site, when I saw this site mentioned in Barron's this week and decided to give it a try. My programming skills are rusty at best, so I need some help.

Here is the strategy. I don't remember where i saw it, but I wrote it down and am writing it from my notes. I wish I had the source written so I could give credit to where credit is due.

I am using two sectors to be held. I will list the ETFs to be used below. There are a few overlaps in the ETFs (eg. healthcare and biotech - IYH and IBB), but if a sector is doing very well, then it maybe worth just piling on to it.

Please note, ROC = rate of change = [(Close - Close n periods ago) / (Close n periods ago)] * 100
X= Delay Period, which hopefully means will avoid sectors that are peaking.
For this case, I am looking at 120 day ROC, but for buying purposes, I will look at the 120 day ROC of ETFs 20 days ago.

The strategy is long only and is simply stated as-
1. At the end of each month, RANK the group of ETFs based on their ROC(n=120 Day), X(20 days back).
2. At the open of the 1st trading day of each month, BUY the Top 2 ETFs, except if they closed below their 120 day moving average. In such case, buy TLT (long-term Treasury bonds) instead.
3. Repeat every month.

ETF universe consists of the following:
TLT
IYC
ITB
IYK
IYE
IYF
IBB
IYH
IYJ
IYW
IYM
IYZ
IDU
ICLN
IGF
IYR
IFGL

Thank you for your help.
Maji

import pandas def lastbday(dtime): '''Check if date is the last business day of the month''' return pandas.bdate_range(dtime.strftime('%m/%d/%Y'), periods=1, freq='BM')[0].date() == dtime.date() def roc(data, n, x): '''Calculate rate of change''' # FIXME: to be implemented pass def rank(data, etfs): '''Calculate rank for the set of etfs.''' # FIXME: this is a fake implementation import random return [(random.randrange(-100, 100), etf) for etf in etfs] def initialize(context): context.etfs = [sid(23921), sid(21644), sid(32011), sid(21650), sid(21646), sid(21522), sid(22445), sid(21648), sid(21649), sid(21524), sid(21651), sid(21638), sid(36480), sid(44539), sid(21652), sid(35095)] context.tlt = sid(23921) def handle_data(context, data): if lastbday(get_datetime()): (rank1, topetf1), (rank2, topetf2) = sorted(rank(data, context.etfs))[-2:] cash = context.portfolio.cash if data[topetf1].price > data[topetf1].mavg(120) and \ data[topetf2].price > data[topetf2].mavg(120): for etf in topetf1, topetf2: order(etf, amount=cash/2/data[etf].price) log.debug('bought %s' % etf.symbol) else order(context.tlt, cash/data[context.tlt].price) log.debug('bought TLT')