The saying goes that you’re better off keeping your mouth closed and looking like a fool than opening it and removing all doubt. Trading reminded us again about the wisdom in those words.

We’d warned that markets showed excessive arbitrage. Arbs capture net spreads between opposing trades and care little about price appreciation. When it’s high in the broad markets (and in specific issues, too), it often points to impending switches in the direction of money, say from one mix of assets to another. Why? Traders apparently detect algorithmic activity and move to profit from it.

We’d seen the same thing in June and thought it might occur again. It didn’t, and the market removed all doubt that we were the fools we appeared to be.

Or so it seems. Are investors responsible for this nice recent run back to positive turf on the major market measures? We can only share general observations from data. The data are what they are. We observed two developments on 7/22. First, Nomura traded more than half our client base that day, a rarity. Nomura had become a primary dealer to the Spanish Treasury a few days earlier, and it’s also a primary dealer for the US Treasury. It owns the trading platform Instinet and Lehman’s Asian and European operations.

How do those facts bear on equity trading when treasuries are a credit market? We can only speculate. But we believe banks can effect carry trades – borrowing and paying interest to earn higher interest on something else.

At the same time Nomura rippled through equity markets, we saw a surge in conventional program trading. Our measures differ from what the exchanges use. Different kinds mean different things. In this case, it was the biggest firms helping global institutions manage transitions from one asset class to another. While market volumes have been relatively light, this increase in market share by big program traders attracted a surge in mathematical trading, which identifies disparities in market structure and capitalizes on them.

Together these could represent a bait-and-switch, not fundamental investment. Banks with trading technologies and commitments to governments to help them fund operations could get money from overnight treasury or central-bank facilities at low cost and use it to trade in equity markets and attract demand. Profits from these operations provide funds for use in required bidding at government primary-dealer auctions. It also gives the bond market a head feint by pulling demand to equities to improve bond-market rates and prices.

We’re not saying this happened. But trading data show that confusion reigns in equity trading markets. That’s not a mark of rational investment. Money may respond sporadically to earnings, and we do certainly see that. But the broad behavior is a mix of discordant sentiments and reactions. We see that, too. Somebody buys in a dark pool, and programs react to it, and algorithms follow, and speculative traders and all the auto-quote systems (translation: “high frequency trading”) tag along, and none of it knows the worth or driver.

Following rules of deductive reasoning, we conclude that if it’s not rational, it’s something else. What else would it be? Well, what dominates the global financial agenda now? Government debt.

So be ye not lulled into a false sense of security, IR pros. Trouble lurks below the rapids. But for now, we’re rafting on big programs, which seem, right or wrong, to know more than you and me.