I want to minimize the drawdown on a strategy. What is the best to go about it ? I have tried using stop loss, but that brings down my Sharpe significantly. How do you guys go about managing risk on your ETF/ Stock strategies?

I`d try to add options to the strategy to "smooth" it

What kind of strategy ? Single stock or Multi stock ? Single stock you get drawdowns because your errors are become autocorrelated, ie you are continuing to guess wrong. It could mean that your forecaster has been thrown by a fat tail or something or the regime has suddenly changed, volatility, liquidity change etc. Multi stock as above, but it may also mean that you have some correlated variables in your strat so the effect is compounded. So say you have an ETF on say oil stocks, if your forecast is done separately on each stock, and then you spit out all your forecasts and then all say to go long ..... and get it wrong as in the single stock case then you'll compound the drawdown....In reality, it's a bit more complicated than that but should point you in right direction. Usually unless part of the strat, avoid putting sticky plasters on a strategy in the form of stop losses, stock cuts, options etc. Not saying they can't be used, but often they are used to desperately try and make strat look better when the focus should be on improving the model, which is the tough bit imo and you generally will be finding models that don't work over those that do.

single stock strat... yes the aim should be to improve the signals but at the same time I dont want my short or long positions to against me for too long. Which is why I need a mechanism to cut losses even thought if I held onto that position I would eventually have made money. Think about it this way, I want to minimize the number of margin calls.

If you don't want to have trouble with margin calls .... then don't trade commodites on the CME ... but I jest .... Well the stop loss .... has to be a function of your strat then. So in a simple example if you are trend following you might use a lower bollinger band to get out of a long position. However in general I'd say stop losses are bad as often the market looks for these, will go through them taking your position out and then move in your favour. As Rweinsh said maybe buying some OTM options would give you the insurance you need. But of course those cost theta. So again it comes down to your signal ... if you have a position that is long ... ie you think something is cheap ... if it moves down then is it not even cheaper ? Maybe worth getting in a bit more then ? Or has something changed and it's now a flat to short ? Either your signal says this or doesn't ? Which is why I believe it's better to have the signal tell you what to do, rather than having a stop loss. Why don't you try to analyse why stop's lower your sharpe ? Is that because that's when the profitable trades arise ? Just after it's gone through your stop and whereas previously you'd then catch the wave up, now you have no position due to being stopped out ?

This is slightly echoing Blades final paragraph of first post, but I'd think of looking at ways you optimised your model. If you're struggling to reduce drawdowns to an acceptable level then look back at how you optimised, can you optimise with greater efficiency? You're idea is sound if it makes money in the long run but you might need to reverse your work flow and try to tackle what you've learned to be the end problem earlier in the model. If you're not already, increase the information you gain from each trade on a backtest, max drawdown, max profit etc. Doing so should again help you to analyse how good the strategy is and if you're holding trades too long, along with indicating optimum levels for stops and limits. From my own experience its much harder working on code/decisions to close trades than open so you need to know as much as possible about what happens to the price/pnl when your model triggers open orders in order to help you work on closes."I need a mechanism to cut losses even thought if I held onto that position I would eventually have made money" But how do you know you'd have eventually made money, no body ever can! Don't be emotional with your strategy.