i priced the synthetic stcdo tranche using gaussian copula monte carlo simulation.to find delta of the tranche i increase the spreads of all obligators by 1 bp and calculate the change in spread of the tranche.repeat the same procedure by decreasing cds spreads of all obligators by 1.tranche delta is the average of two changes.Suppose notional is $100 million.no of obligators- 125equal weights of each obligator ie 0.8 millionspread of tranche is 120 bp.and delta of tranche is 6x.so for hedging i will sel 600 million in total and if i sell equal amounts of each obligator i sell $4.8.but that is wrong as some credit might be more sensitive.so i do the foolowing thing.say the obligator 1 is albertson. we will now calculate the delta of albertson. suppoe cds spread is 48 bp.i increase its spread to 49bp.we find the spread of tranche again suppose it comes out to be 120.05 bps.the payments in the deal would have been-year 1 2 3 4 5 1.20% 1.20% 1.20% 1.20% 1.20%+100millionif we discount this by 1.20% we will get back 100.but we discount this by 1.205% suppose we get 99.98.now consider the albertson cds.with our initial assumption of same sensitivity of all obligators we sell $ 4.8 million=600/125payments involved in this dealyear 1 2 3 4 5 0.48% 0.48% 0.48% 0.48% 0.48%+4.8 millionnow qw discount this by 0.49% and suppose we get 4.7.so delta of albertson = (100-99.98)/(4.8-4.7)in the revised strategy we 0.8 million * delta of albertson.is this hedging strategy justified and practised?can any one give me literature on it?

You could take a look at the following articles:ML Correlation tradingSG Correlation researchHope it helps

reply thanks for the method. my problem is solved.