I promised above somewhere that I'd report back on my experiences with Gunbot. I've been into Bitcoin for about a year, with only a little activity. Having learnt about lending bots, I have been lending for a few months.
Having learnt about trading bots I was intrigued, read through this complete thread, and tried out Gunthar's trading strategy (first post) manually - and it worked! I ordered Gunbot about 3 weeks ago. I set it up quickly as I was about to go on holiday. I did it first on my own PC, leaving it running overnight. Then transferred it to a VPS with help from Captain Corporate. I have only used the BB strategy, and with Gunthar's default settings, as I don't yet feel confident trying anything else.
I kicked off the bot seriously on 13 August with about 20 coin pairs, and added some more a few days later. I then went on holiday. I monitored it while on holiday and noticed it was building up a bunch of bags. I got nervous, and turned off most pairs after about 10 days, just leaving those with bags running in the hope they would be sold. Within the next week it sold off all the bags successfully, so I probably needn't have worried. Apart from one bag where the altcoin had dropped significantly - I sold it manually because I didn't want to keep it indefinitely, and although I lost some money on that sell it was less than I had made previously on that pair so overall that pair was still profitable.
Here's a screenshot from my Excel spreadsheet for the period. It's in BTC, and also in British pounds as I'm in the UK (one pound is $1.3).
I paid 0.09BTC for the Gunbot. Over the period I made 0.15BTC so I'm happy. I will definitely be continuing to use it - I've just kicked it off again on the top pairs in the list above (by %profit per day, so LTC, FCT, XEM, XRP, etc.). I'm avoiding any that hung around in bags for more than a week (BTCD, BURST, DBG, etc.). The %profit and %profit/day are the key metrics as far as I'm concerned. You should largely ignore the actual profit amount as that depended entirely on how much I risked in a trade. I started with trades of 0.003, then 0.01, and now use 0.1, so different coins in the table will have trades at different amounts. The %profit is what actually matters.
It has been an interesting and exciting experiment. My hat goes off to Gunthar, and I recommend Gunbot (with the usual qualifications about learning what you need to know, not risking what you can't afford to lose, etc.).