The defendants, partners of the stock brokerage house of Jacques Coe & Co. appeal from a judgment, rendered after a non-jury trial, in favor of its customer, James Wood General Trading Establishment, a corporation of the Principality of Liechtenstein, in the amount of $34,-486.28. The complaint is framed in a double aspect of contract and tort, and is based upon the simple charge that the customer placed a G. T. C. (good until cancelled) order with the brokers to sell 1000 shares of Baltimore & Ohio common stock, then in the custody of the brokers, “when the stock could be sold for more than $58 per share,” that the stock sold at prices in excess of 58 on July 25 and 26, 1957, that the brokers failed to sell the stock, and that, as the brokers failed to remind the customer of the G. T. C. order, and as the customer had forgotten about it, his damages should be calculated at the difference between the price at which the stock could have been sold “above 58” and the mean *653selling price of the stock for the week following the time the customer discovered or finally remembered the G. T. C. order. By that time the stock had taken a precipitous drop. The trial judge took this view of the case, found that both parties had forgotten about the G. T. C. order, and directed judgment against the brokers for the full amount claimed. There is no claim in the complaint that the brokers were otherwise guilty of any breach of their fiduciary or other duties to the customer, or that they failed properly to exercise any power or authority conferred upon them. The failure to sell was, at the trial, ascribed by the customer to the fact that the brokers had also forgotten about the alleged G. T. C. order. The opinion is published at 191 F.Supp. 330.
We find no order to sell the stock, but the granting by the customer of a mere authority or power to sell at “above 58” according to the best judgment or discretion of the brokers. Moreover, we think the finding that both parties forgot about “the order” must be set aside as largely based upon the assumption that a G. T. C. order to sell had been given and upon other misconstructions of the proofs. As we proceed to state the case as disclosed by the evidence, it will appear that the customer was at all times aware of the progress of the market, and that by an exchange of letters of August 5 and 12, 1957, the dealings between the parties had crystallized into a G. T. C. order to sell 500 shares of the B & 0 stock at 61, and not to sell the other 500 shares, so that in any event, even assuming arguendo that the alleged G. T. C. order was given in November, 1956, and that it was a breach of contract or negligence on the part of the brokers not to sell when the price of the stock rose above 58 on July 25 and 26, 1957, and further assuming arguendo that both parties had forgotten about the alleged original order, the customer's damages should have been limited to the difference between what the stock could have been sold for on July 25 and 26,1957 and the lowest price to which it fell prior to the new arrangement. It is also clear to us that, even if the brokers had forgotten their instructions and had wholly failed to exercise any judgment or discretion in the matter, there is still lacking any proof that, had they remembered the instructions, and exercised their judgment and discretion, they would have sold the stock on July 25 or 26, 1957. Accordingly, we have concluded that there is no point in remanding the case to permit plaintiff to restate its claim and on a new trial litigate its assertion, in the brief filed with us, that it may be entitled to damages for the failure of the brokers to exercise any discretion that might have been conferred upon them. As the customer suffered no provable damage the complaint must be dismissed.
Jurisdiction is based upon diversity and New York law governs the case, as seems to be conceded by the parties.
The representatives of the parties were Mario Quarti, a resident of Italy, agent of the customer, and Emilio Mauricio Moretti, the brokers’ customers’ man, located in New York City. They were old friends and had long done business together, as Quarti, on his own account or representing the Liechtenstein corporation, had conducted a series of transactions in the purchase and sale of stocks for profit. The correspondence between these two old friends jvas in Italian, but the translations received in evidence are stipulated to be in all respects accurate.
In October, 1956 the customer was long 1000 shares of B & O, 4500 shares of Menasco Manufacturing Co., and 500 shares of Vanadium Corporation, all of which stocks were in the custody of the brokers. Moretti made certain recommendations to Quarti concerning the sale of these stocks, and Quarti replied by a letter that has been lost.1 Moretti’s re*654ply, by letter of November 19, 1956 repeats the instructions given by Quarti, however, and both sides rely upon it as correctly reflecting the substance of these instructions. The relevant portions of this letter control the case, and they are as follows:
“BO. I enclose two clippings of news published today on BO. As you see, I have not erred too much in predicting that the dividend, which had to be declared these dáys, would have been probably a little higher than the one of $2.00 of last year. It was in fact $2.50, 50 cents more than last year. At the same time they have declared for next year 1957, four dividends of 50 cents each to be paid quarterly, instead of all in one time. Earnings also have been better; for the first ten months of 1956, about 26 million versus about 21 million for the equivalent period of last year. Result: the stock, which Friday had closed at 52%, went down today to 48%. To this might have contributed the general tone of the market which seems to have been hit today by rumors of possible new hostilities in the Middle East etc. / /1 note that you authorize to sell this stock above 58. Unfortunately, I doubt that it will go there soon; but with time I imagine it might be able to do it.
Vanadium. I enclose credit note for a dividend (net $175). The stock, which is still doing very well, as far as earnings are concerned, is now 46. Thank you for the authority to sell it above 49. Except in case of a market crash, I think it could go there soon.
Menasco. Stays still around 5%-5%. It is alright to sell it at 6% or better. Being to some good extent the beneficiary of our defensive preparations (airplanes) I think it will go there, maybe rather soon.” (Underscoring as in letter.)
Upon the receipt of these instructions the brokers did not proceed to place orders on the floor of the New York Stock Exchange to sell any particular number of shares or to sell any shares whatever of the customer’s B & O, Vanadium and Menasco stocks, and of course no monthly confirmations of such orders were sent to the customer as there were no such orders to confirm. In December, 1956 Vanadium broke 49 and the 500 shares were sold by the brokers in 100 and 200 share blocks at prices ranging from 49% to 50%; notations of these sales were sent and the receipt thereof acknowledged. In January, 1957 Menasco broke 6% and the 4500 shares of Menasco were sold by the brokers in 100, 200, 300 and 600 share blocks at prices ranging from 6% to 6%, notations of these sales were also sent to and received by the customer.
In his letter of November 19, 1956, quoted above, Moretti noted the authority to sell B & O “above 58,” and he added “I doubt it will go there soon.” He proved to be right about this, and B & O remained throughout the first six months of the year 1957 just about where it had been in the previous November, or lower. There is no dispute about the fact that both Quarti and Moretti were watching the market quotations of B & O and Quarti knew that the stock had not been sold and that it had not reached 58.
On July 16, 1957 Moretti wrote to Quarti as follows:
“ * * * BO goes well, and is now 55. There is a possibility— only a possibility — that they may increase the dividend, which is 50 cents quarterly, to perhaps 60 or 70, or that they may give, towards November, an extra dividend of $1.00. My opinion in this respect is that it would be well to sell around 58-60. But, since I remember that you received the information, when you bought it, from London, I think it would be well that you should consult again London to know what they think now of BO, and at what price they would sell it. * * * ” (Underscoring as in letter.)
This letter was received by Quarti prior to July 25 and 26,1957 when B & O rose above 58, and Quarti was aware of the fact that there were sales on those two days at prices ranging between 58% and 58%. The trial judge properly found that the customer’s 1000 shares of B & O *655could have been sold on those two days at 58Vs. But Quarti expressed no dissatisfaction and sent no further instructions until August 5, 1957. The reason is apparent on the face of Quarti’s letter of August 5,1957. He expected the stock to go higher; he had received favorable news from his correspondent in London, “all roses and flowers”; he admitted he was tempted to cable to sell at 58, but he is now thinking in terms of “above 60” or “let us say 60 net for me.”
As soon as Moretti received Quarti’s letter of August 5, 1957, he wrote on August 12, 1957 accepting Quarti's “idea,” and the next day an order was placed G. T. C. to sell 500 shares at 61, which would net the customer “something more than 60.” The only inference to be drawn as to the remaining 50G shares is that they were to be held to be sold later at a price higher than 61, if it could be obtained. Monthly confirmations of this G. T. C. order were sent, Quarti acceded to the new plan of operations, and all was well— until B & O went down.
When it had reached the thirties Quarti began looking around for someone to blame, and he found the letter of November 19, 1956 quoted at the beginning of our recital of the facts. In the course of time Quarti, despite the fact that he was at all times aware of the ups and downs of B & O and in regular correspondence with Moretti, evolved the claim that “the original order” given in November, 1956 was a G. T. C. order to sell when B & O got “above 58”; that the brokers were at fault in not “reminding” him that such an order had been given, that he had accordingly forgotten all about it; that the brokers committed a breach of contract or were guilty of negligence in not selling the stock on July 25 or July 26, 1957 at 58%; that he learned of this breach of contract or tortious conduct only on November 14, 1957, when he “found” Moretti’s letter of November 19, 1956, and when B & O had gone down to about 29%. The result has thus far been a judgment against the brokers for $34,486.28. This judgment must be reversed.
I
There was no order to sell the B & O “above 58,” and, consequently, no G. T. C. order to sell. For the same reason there was no occasion to place any orders to sell on the floor of the New York Stock Exchange, or to send confirmations of such orders to the customer. Nor is there anything to the contrary in the testimony of the expert called by defendants to describe the customs and practices of the New York Stock Exchange. The confirmations he was talking about related to G. T. C. orders to sell, and, as the inquiries concerning G. T. C. orders proceeded, one of them related to a G. T. C. order to sell when the stock reached a certain level. When the witness characterized such an order as “discretionary,” the context of his answer indicates that he was referring to the discretion involved in selling a block of stock pursuant to a G. T. C. order, “a little at a time,” as the market rises. Nothing in the testimony of this witness would sustain a finding that it was the custom and practice of members of the New York Stock Exchange to remind a customer from time to time that the customer had given the brokers authority in their discretion to sell a certain stock when, as and if it reached a certain level.
True it is that the instructions of the customer relative to the B & O stock were the same as those concerning the Menasco and Vanadium stocks. But there were no G. T. C. orders to sell any of these stocks and, upon the receipt of the instructions from the customer, no orders to sell were placed on the floor of the New York Stock Exchange, nor were any monthly confirmations sent although there was a sufficient interval prior to the sales of the Menasco and Vanadium stocks to send confirmations, as would have been required in the case of G. T. C. orders. What the brokers did, without the receipt of further instructions from the customer, was to wait until these stocks reached the price levels mentioned in the instructions, and then they were sold because in the exercise of the judgment and discretion of the brokers it was a good *656time to sell them at the prices and in the small blocks at which they were sold.
Nor do we think it of any significance that, in authorizing the brokers to buy 500 shares of a certain other stock in a letter of June 4, 1957, Quarti specifically wrote that he was giving them “discretion” to buy the stock.
What we have in this case is not an order to sell but the conferring by a principal upon his agents of authority in their discretion and according to their best judgment to sell 1000 shares of B & O “above 58.” Findings below to the contrary are clearly erroneous and they are reversed and set aside.
II
Passing for the moment the rather confused question of whether either or both Quarti and Moretti forgot about the instructions given in November, 1956, and the allied question of what damages might be deemed to flow from a failure by the agent to exercise the discretion and judgment that he was authorized to exercise, and impliedly agreed to exercise, we turn to a consideration of the effect of the letters of August 5 and 12, 1957.
Even if we assume arguendo that the brokers had committed some wrong by way of breach of contract or otherwise by failing to sell the B & O stock at 58% on July 25 or 26, 1957, these letters and the conduct of the parties immediately thereafter resulted in a new contractual arrangement that precluded the possibility of a sale by the brokers of any of the B & O stock at less than 61. Thus by its own voluntary acts the customer limited the recoverable damages to the difference between the price at which the stock could have been sold on July 25 or 26, 1957, 58%, and the value of the stock computed upon the basis of the price most favorable to the customer- in the interval of time between July 25 or 26, 1957 and August 18, 1957 when the order to sell 500 shares at 61 was given and confirmed. The New York cases are very clear on this point.
Speyer v. Colgate, Gen. T. 1st Dep’t, 1875, 67 Barb. 192, held that when through a broker’s fault a sale is not made, if the customer later advises the broker not to sell the property involved, from that time on any decline in price is at the customer’s risk. In this case, as a result of the broker’s failure to deliver the customer’s gold it was not sold. After the customer discovered this he directed the broker not to sell the gold, but to wait for further advice. As held by the Court at p. 200:
“The market was then falling, and continued to fall. After that, as he had directed the defendants not to sell, but to wait for advice, they held the gold remaining on hand under his orders, and consequently at his risk. He could have at once relieved himself from that risk by directing that the gold should be sold; but he did not do it, and the loss from that time resulted from this omission.”
While it is true that in Speyer v. Colgate the customer knew that his order had not been carried out and in our case the trial judge found that Quarti had forgotten his prior instruction at the time the August arrangement was made, this difference does not affect the applicability of the New York rule. The significant fact is that as a result of the new arrangement, made pursuant to the customer’s suggestion, the broker could no longer sell the B & O merely “above 58.” Furthermore, beyond doubt, the customer’s acts lulled the broker into a sense of security, to the effect that he acted properly in not selling on July 25 or 26, and thereby prevented him from reducing the damage. Any subsequent decline in price therefore was at the risk of the customer. Craig v. Pierce, 1st Dep’t, 1930, 231 App.Div. 159, 246 N.Y.S. 573; Hope v. Lawrence, N. Y. Gr. Tr., 1867, 50 Barb. 258; cf. Hirsch v. Deschler, 1st Dep’t, 1931, 231 App.Div. 237, 247 N.Y. S. 143, aff’d 1931, 257 N.Y. 570, 178 N. E. 799.
The reason none of these cases was drawn to our attention by counsel is that *657counsel misconceived the effect of the new arrangement to sell 500 shares at 61 and not sell the other 500 shares. The argument made in the brokers’ pleading as well as at the trial was that the new arrangement above referred to constituted a waiver, or an estoppel, or a ratification of the brokers’ failure to sell “above 58.” But the trial judge properly held that none of these defenses could apply if Quarti had no knowledge or recollection of the instructions he gave in November, 1956. The defense of account stated practically fell of its own weight.
As we look at this phase of the case it is wholly immaterial whether either or both parties had forgotten about the instructions given in November, 1956. We cut right through to the question of damages. The new contractual arrangement was voluntarily entered into, it was in all respects valid and binding upon both parties, and it interposed an insuperable obstacle to the taking of any action by the brokers to sell the stock before it went any lower and thus to reduce the amount of any obligation to the customer for failure to sell the stock at 58%. A claimant acts at his peril if by his own voluntary acts he creates a situation in which the responsible party is not only led to believe that he did no wrong, but also is duty bound not to take the only action whereby he could prevent the damages from piling up. This principle is particularly applicable to transactions in which one of the parties may be disposed to play a heads I win tails you lose game.
Ill
There remains for discussion a confused and troublesome feature of the case. While no such claim was asserted in the complaint or at the trial, it is argued in appellee’s brief, as we read it, that, even if the order was “discretionary” the customer was entitled to an award of damages, because Moretti had forgotten about “the order,” and exercised no discretion whatever.
We have already set aside the finding that both Quarti and Moretti had forgotten about the instructions given by Quarti in November, 1956, because the finding is based upon a misconception of the basic facts of the case. Indeed, there is much to indicate that the instructions were not forgotten. Quarti was not produced at the trial and he gave no testimony. Moretti’s deposition was taken by plaintiff and certain more or less disconnected portions of it were read into evidence by both parties. But Moretti was not called as a witness at the trial, although it would seem that the brokers could have called him had they chosen to do so. In any event, a portion of Moretti’s deposition, as read at the trial, is to the effect that he had not forgotten the instructions; and his repeated reports to Quarti about the B & 0 prospects over the period intervening between November, 1956 and July, 1957, seem consistent with his position that he did not want to place an order to sell while reports concerning B & 0 were so favorable and Quarti was so situated that he could cable specific instructions to sell at any time he wished to do so.
The statement in appellee’s brief, “Moretti concedes that he failed to exercise any discretion concerning the desirability of a sale,” is based, we think, upon a misconstruction of his testimony. He did admit that when B & 0 rose above 58 for the two days in July, 1957 he did not place any orders to sell. The questioning proceeded on the subject of whether he could have disposed of the stock had he placed an order to sell at 58%. It was in this context of what the market would have absorbed on those days and at what prices, that we find the following question and answer: “Q. You did not use this discretion, as you describe it, at all? A. No.” We think all Moretti meant by this is that he could have placed orders to sell in various amounts of shares and at various prices, but he did not use this sort of discretion at all, as he was in touch with Quarti by letter, and he thought the stock was going up. He also testified: “If I had proceeded to put the order to sell, I might have sold, I don’t know how much. The market at that particular moment was very strong. I *658might have decided to place orders at 59, 59 and a half, something of the kind, trying to do the best for him, of course, attempting to get the better price for him since the market looked very strong at that particular time.”
Under these circumstances Moretti decided to wait for a reply by Quarti to his letter of July 16, 1957. And when that reply arrived it appeared that Quarti had been tempted to cable to sell at 58, but did not do so because of the “all roses and flowers” report on B & 0 he had received from his London informant. In other words, Quarti also thought the market was strong and that B & 0 would go up.
Despite all this, and after reading all the correspondence in evidence, it remains true that, subsequent to the original acknowledgement of the instructions, and prior to the time the controversy between the parties arose, neither Moretti nor Quarti made any specific reference in any of their letters to the instructions given in November, 1956. Their conduct is just as consistent with the supposition that they remembered it as it is with the supposition that they had forgotten about it.
Having considered the instructions as a G. T. C. order to sell “above 58,” the trial judge of course made no findings on the subject of the exercise of discretion. Nor can we say what finding he would have made on the subject of the state of mind of the participants relative to the instructions as given, had the trial judge viewed the facts in the light described in this opinion.
However, even assuming Moretti had forgotten the instructions as given, and that he exercised no discretion whatever, it is difficult to find in this record any evidence of damage. What proof can there be that, if Moretti had not forgotten the instructions, he would have sold the stock on July 25 or 26, 1957 at 58% or any other price ? How can such a conclusion be arrived at other than by the merest guesswork and speculation ? The plain fact is that, in the circumstances of this case there is and can be no evidence of any probative value to support a finding that he would have sold. On the contrary, there is substantial evidence to support an inference that he would not have sold the B & 0 at the prices prevailing in the market on July 25 and 26, 1957.
Damages cannot be based on speculation. The difficulty of measuring and appraising the amount, once it is clear that actual damage has been sustained, is another matter. WTiere, as here, the existence of any damage whatever would have to be a matter of guesswork, no recovery is allowed. Broadway Photoplay Co. v. World Film Corp., 1919, 225 N.Y. 104, 121 N.E. 756; Steitz v. Gifford, 1939, 280 N.Y. 15, 20, 19 N.E.2d 661, 664, 122 A.L.R. 292 (dictum); Wakeman v. Wheeler & Wilson Mfg. Co., 1886, 101 N.Y. 205, 209, 4 N.E. 264, 266 (dictum); 2 Harper & James, Torts § 15.2(3), at 878 (1956) ; 5 Williston, Contracts, § 1345 (rev. ed. 1937); see Restatement (Second), Agency § 400, comment b (1958).
Where actual damage depends on an exercise of discretion, and there is no evidence from which it can reasonably be determined how such discretion would be exercised, no damages are recoverable. 25 Fifth Ave. Management Co. v. Clark, 1952, 280 App.Div. 205, 112 N.Y.S.2d 117, aff’d 304 N.Y. 808, 109 N.E.2d 469.
Reversed and complaint dismissed with costs.
. Moretti testified that Quarti’s letter of November 1956, together with the rest of the file on Moretti’s 1956 correspondence relative to appellee’s account, was placed in storage some time between March and June, 1957, and that a search had failed to locate this file.